project
Hakan Ekiz/Batuhan Tunçel/Yusuf Beşiroğlu - IE48B - Fall2021
24 January 2022
require(data.table)## Loading required package: data.tablerequire(ggplot2)## Loading required package: ggplot2require(repr)## Loading required package: reprrequire(rpart)## Loading required package: rpartrequire(rattle)## Loading required package: rattle## Loading required package: tibble## Loading required package: bitops## Rattle: A free graphical interface for data science with R.
## Version 5.4.0 Copyright (c) 2006-2020 Togaware Pty Ltd.
## Type 'rattle()' to shake, rattle, and roll your data.require(TSrepr)## Loading required package: TSreprrequire(TSdist)## Loading required package: TSdist## Loading required package: proxy##
## Attaching package: 'proxy'## The following objects are masked from 'package:stats':
##
## as.dist, dist## The following object is masked from 'package:base':
##
## as.matrix## Error: package or namespace load failed for 'TSdist' in loadNamespace(i, c(lib.loc, .libPaths()), versionCheck = vI[[i]]):
## there is no package called 'rgl'require(dtw)## Loading required package: dtw## Loaded dtw v1.22-3. See ?dtw for help, citation("dtw") for use in publication.require(zoo)## Loading required package: zoo##
## Attaching package: 'zoo'## The following objects are masked from 'package:base':
##
## as.Date, as.Date.numericrequire(genlasso)## Loading required package: genlasso## Loading required package: Matrix##
## Attaching package: 'Matrix'## The following object is masked from 'package:bitops':
##
## %&%require(e1071)## Loading required package: e1071require(lubridate)## Loading required package: lubridate##
## Attaching package: 'lubridate'## The following objects are masked from 'package:data.table':
##
## hour, isoweek, mday, minute, month, quarter, second, wday, week,
## yday, year## The following objects are masked from 'package:base':
##
## date, intersect, setdiff, unionrequire(forecast)## Loading required package: forecast## Registered S3 method overwritten by 'quantmod':
## method from
## as.zoo.data.frame zoorequire(dplyr)## Loading required package: dplyr##
## Attaching package: 'dplyr'## The following objects are masked from 'package:data.table':
##
## between, first, last## The following objects are masked from 'package:stats':
##
## filter, lag## The following objects are masked from 'package:base':
##
## intersect, setdiff, setequal, unionoptions(repr.plot.width=15, repr.plot.height=8)
current_folder=getwd()The latest input
dat=fread('bulk_imbalance.csv')
weat=fread('2022-01-21_weather.csv')
head(dat)## date hour net upRegulationZeroCoded upRegulationOneCoded
## 1: 2019-01-01 0 -1934.795 0 0
## 2: 2019-01-01 1 -750.866 0 0
## 3: 2019-01-01 2 -982.767 0 0
## 4: 2019-01-01 3 -764.620 0 0
## 5: 2019-01-01 4 -1233.289 0 0
## 6: 2019-01-01 5 -1469.061 0 0
## upRegulationTwoCoded downRegulationZeroCoded downRegulationOneCoded
## 1: 0 2288.567 0
## 2: 0 808.083 0
## 3: 0 1060.750 0
## 4: 0 1120.483 0
## 5: 0 1874.633 56
## 6: 0 2125.000 56
## downRegulationTwoCoded upRegulationDelivered downRegulationDelivered
## 1: 0 0 -1934.795
## 2: 0 0 -750.866
## 3: 0 0 -982.767
## 4: 0 0 -764.620
## 5: 0 0 -1233.289
## 6: 0 0 -1469.061
## system_direction
## 1: Negative
## 2: Negative
## 3: Negative
## 4: Negative
## 5: Negative
## 6: Negativehead(weat)## date hour lat lon variable value
## 1: 2018-12-31 0 36.5 32.5 DSWRF_surface 0
## 2: 2018-12-31 1 36.5 32.5 DSWRF_surface 0
## 3: 2018-12-31 2 36.5 32.5 DSWRF_surface 0
## 4: 2018-12-31 3 36.5 32.5 DSWRF_surface 0
## 5: 2018-12-31 4 36.5 32.5 DSWRF_surface 0
## 6: 2018-12-31 5 36.5 32.5 DSWRF_surface 0c1 = dat
c1 = c1[,list(date,hour,net,system_direction)]
c1[,datetime:=ymd(date)+dhours(hour)]
c1=c1[order(datetime)]
c2 = c1[,list(datetime,net,system_direction)]
c2## datetime net system_direction
## 1: 2019-01-01 00:00:00 -1934.795 Negative
## 2: 2019-01-01 01:00:00 -750.866 Negative
## 3: 2019-01-01 02:00:00 -982.767 Negative
## 4: 2019-01-01 03:00:00 -764.620 Negative
## 5: 2019-01-01 04:00:00 -1233.289 Negative
## ---
## 26804: 2022-01-21 19:00:00 0.000
## 26805: 2022-01-21 20:00:00 0.000
## 26806: 2022-01-21 21:00:00 0.000
## 26807: 2022-01-21 22:00:00 0.000
## 26808: 2022-01-21 23:00:00 0.000w1 = weat
w1[,datetime:=ymd(date)+dhours(hour)]
w1=w1[order(datetime)]
w1## date hour lat lon variable value datetime
## 1: 2018-12-31 0 36.50 32.50 DSWRF_surface 0 2018-12-31 00:00:00
## 2: 2018-12-31 0 37.00 35.50 DSWRF_surface 0 2018-12-31 00:00:00
## 3: 2018-12-31 0 38.00 32.50 DSWRF_surface 0 2018-12-31 00:00:00
## 4: 2018-12-31 0 38.50 27.00 DSWRF_surface 0 2018-12-31 00:00:00
## 5: 2018-12-31 0 39.75 30.50 DSWRF_surface 0 2018-12-31 00:00:00
## ---
## 1129964: 2022-01-24 23 38.00 32.50 wdir_10m NA 2022-01-24 23:00:00
## 1129965: 2022-01-24 23 38.50 27.00 wdir_10m NA 2022-01-24 23:00:00
## 1129966: 2022-01-24 23 39.75 30.50 wdir_10m NA 2022-01-24 23:00:00
## 1129967: 2022-01-24 23 40.00 33.00 wdir_10m NA 2022-01-24 23:00:00
## 1129968: 2022-01-24 23 41.00 28.75 wdir_10m NA 2022-01-24 23:00:00Ordering by date-time for all locations and their weather conditions
Also bringing all parameters together
w1 = w1 %>%
mutate(loc =
case_when(lat <= 36.50 ~ "loc1" ,
lat==37.00 ~ "loc2" ,
lat==38.00 ~ "loc3" ,
lat==38.50 ~ "loc4" ,
lat==39.75 ~ "loc5" ,
lat==40.00 ~ "loc6" ,
lat>=41.00 ~ "loc7" )
)
w1= as.data.table(w1)
w1## date hour lat lon variable value datetime
## 1: 2018-12-31 0 36.50 32.50 DSWRF_surface 0 2018-12-31 00:00:00
## 2: 2018-12-31 0 37.00 35.50 DSWRF_surface 0 2018-12-31 00:00:00
## 3: 2018-12-31 0 38.00 32.50 DSWRF_surface 0 2018-12-31 00:00:00
## 4: 2018-12-31 0 38.50 27.00 DSWRF_surface 0 2018-12-31 00:00:00
## 5: 2018-12-31 0 39.75 30.50 DSWRF_surface 0 2018-12-31 00:00:00
## ---
## 1129964: 2022-01-24 23 38.00 32.50 wdir_10m NA 2022-01-24 23:00:00
## 1129965: 2022-01-24 23 38.50 27.00 wdir_10m NA 2022-01-24 23:00:00
## 1129966: 2022-01-24 23 39.75 30.50 wdir_10m NA 2022-01-24 23:00:00
## 1129967: 2022-01-24 23 40.00 33.00 wdir_10m NA 2022-01-24 23:00:00
## 1129968: 2022-01-24 23 41.00 28.75 wdir_10m NA 2022-01-24 23:00:00
## loc
## 1: loc1
## 2: loc2
## 3: loc3
## 4: loc4
## 5: loc5
## ---
## 1129964: loc3
## 1129965: loc4
## 1129966: loc5
## 1129967: loc6
## 1129968: loc7w1[,class:=paste(loc,variable)]
w1## date hour lat lon variable value datetime
## 1: 2018-12-31 0 36.50 32.50 DSWRF_surface 0 2018-12-31 00:00:00
## 2: 2018-12-31 0 37.00 35.50 DSWRF_surface 0 2018-12-31 00:00:00
## 3: 2018-12-31 0 38.00 32.50 DSWRF_surface 0 2018-12-31 00:00:00
## 4: 2018-12-31 0 38.50 27.00 DSWRF_surface 0 2018-12-31 00:00:00
## 5: 2018-12-31 0 39.75 30.50 DSWRF_surface 0 2018-12-31 00:00:00
## ---
## 1129964: 2022-01-24 23 38.00 32.50 wdir_10m NA 2022-01-24 23:00:00
## 1129965: 2022-01-24 23 38.50 27.00 wdir_10m NA 2022-01-24 23:00:00
## 1129966: 2022-01-24 23 39.75 30.50 wdir_10m NA 2022-01-24 23:00:00
## 1129967: 2022-01-24 23 40.00 33.00 wdir_10m NA 2022-01-24 23:00:00
## 1129968: 2022-01-24 23 41.00 28.75 wdir_10m NA 2022-01-24 23:00:00
## loc class
## 1: loc1 loc1 DSWRF_surface
## 2: loc2 loc2 DSWRF_surface
## 3: loc3 loc3 DSWRF_surface
## 4: loc4 loc4 DSWRF_surface
## 5: loc5 loc5 DSWRF_surface
## ---
## 1129964: loc3 loc3 wdir_10m
## 1129965: loc4 loc4 wdir_10m
## 1129966: loc5 loc5 wdir_10m
## 1129967: loc6 loc6 wdir_10m
## 1129968: loc7 loc7 wdir_10mw2 = w1[,list(datetime,class,value)]
w2[,class:=as.factor(class)]
w2## datetime class value
## 1: 2018-12-31 00:00:00 loc1 DSWRF_surface 0
## 2: 2018-12-31 00:00:00 loc2 DSWRF_surface 0
## 3: 2018-12-31 00:00:00 loc3 DSWRF_surface 0
## 4: 2018-12-31 00:00:00 loc4 DSWRF_surface 0
## 5: 2018-12-31 00:00:00 loc5 DSWRF_surface 0
## ---
## 1129964: 2022-01-24 23:00:00 loc3 wdir_10m NA
## 1129965: 2022-01-24 23:00:00 loc4 wdir_10m NA
## 1129966: 2022-01-24 23:00:00 loc5 wdir_10m NA
## 1129967: 2022-01-24 23:00:00 loc6 wdir_10m NA
## 1129968: 2022-01-24 23:00:00 loc7 wdir_10m NAtrainw2=dcast(w2[!is.na(value)],datetime~class,value.var='value')
trainw2## datetime loc1 DSWRF_surface loc1 RH_2.m.above.ground
## 1: 2018-12-31 00:00:00 0.00 83.30000
## 2: 2018-12-31 01:00:00 0.00 82.00000
## 3: 2018-12-31 02:00:00 0.00 80.70000
## 4: 2018-12-31 03:00:00 0.00 79.40000
## 5: 2018-12-31 04:00:00 0.00 78.83333
## ---
## 26898: 2022-01-24 17:00:00 348.20 75.60000
## 26899: 2022-01-24 18:00:00 255.46 79.00000
## 26900: 2022-01-24 19:00:00 191.60 78.00000
## 26901: 2022-01-24 20:00:00 153.28 80.00000
## 26902: 2022-01-24 21:00:00 127.74 79.90000
## loc1 TCDC_low.cloud.layer loc1 TMP_2.m.above.ground loc1 wdir_10m
## 1: 0.0 -8.284000 219.0959
## 2: 1.0 -8.104667 217.8237
## 3: 2.0 -7.925333 216.5515
## 4: 3.0 -7.746000 215.2793
## 5: 5.0 -7.995333 216.4846
## ---
## 26898: 3.4 -8.619000 232.9766
## 26899: 2.4 -13.750000 258.6988
## 26900: 1.8 -15.312000 261.6345
## 26901: 1.5 -15.771000 262.4866
## 26902: 1.2 -15.815000 264.9919
## loc1 ws_10m loc2 DSWRF_surface loc2 RH_2.m.above.ground
## 1: 2.790017 0.00 76.1
## 2: 2.725563 0.00 74.6
## 3: 2.661109 0.00 73.1
## 4: 2.596655 0.00 71.6
## 5: 2.551815 0.00 71.3
## ---
## 26898: 1.541491 278.72 69.0
## 26899: 2.451014 197.54 65.9
## 26900: 2.975670 148.14 67.6
## 26901: 3.469357 118.52 73.1
## 26902: 3.329440 98.78 76.8
## loc2 TCDC_low.cloud.layer loc2 TMP_2.m.above.ground loc2 wdir_10m
## 1: 66.00000 4.396000 179.5000
## 2: 59.66667 3.972000 182.0823
## 3: 53.33333 3.548000 184.6647
## 4: 47.00000 3.124000 187.2471
## 5: 31.66667 3.111333 186.7205
## ---
## 26898: 0.00000 3.611000 163.0320
## 26899: 0.00000 1.030000 168.9420
## 26900: 0.00000 0.198000 169.6947
## 26901: 0.00000 -0.651000 168.4363
## 26902: 0.00000 -1.405000 181.3601
## loc2 ws_10m loc3 DSWRF_surface loc3 RH_2.m.above.ground
## 1: 2.3553997 0.00 93.20000
## 2: 2.3094280 0.00 92.93333
## 3: 2.2634562 0.00 92.66667
## 4: 2.2174845 0.00 92.40000
## 5: 2.2073096 0.00 92.06667
## ---
## 26898: 0.3744988 308.36 81.10000
## 26899: 0.3467570 222.70 85.70000
## 26900: 1.1514952 167.02 86.10000
## 26901: 2.1472339 133.62 85.80000
## 26902: 2.8114421 111.34 84.90000
## loc3 TCDC_low.cloud.layer loc3 TMP_2.m.above.ground loc3 wdir_10m
## 1: 95.0 -4.684000 231.8005
## 2: 89.0 -4.884667 245.0924
## 3: 83.0 -5.085333 258.3843
## 4: 77.0 -5.286000 271.6762
## 5: 54.0 -5.995333 217.6031
## ---
## 26898: 79.6 -8.829000 188.7512
## 26899: 77.2 -12.350000 148.3612
## 26900: 74.2 -13.982000 127.2406
## 26901: 72.7 -14.521000 120.5353
## 26902: 73.8 -14.465000 116.5458
## loc3 ws_10m loc4 DSWRF_surface loc4 RH_2.m.above.ground
## 1: 0.8009606 0.00 75.20000
## 2: 0.6506007 0.00 76.03333
## 3: 0.5002407 0.00 76.86667
## 4: 0.3498807 0.00 77.70000
## 5: 0.3796237 0.00 77.60000
## ---
## 26898: 2.4365654 190.82 63.50000
## 26899: 2.6431705 144.14 72.80000
## 26900: 2.8966046 108.96 77.00000
## 26901: 2.8217623 87.18 76.60000
## 26902: 2.6194151 72.64 74.70000
## loc4 TCDC_low.cloud.layer loc4 TMP_2.m.above.ground loc4 wdir_10m
## 1: 0 5.026000 253.8836
## 2: 0 4.732000 257.6694
## 3: 0 4.438000 261.4552
## 4: 0 4.144000 265.2409
## 5: 0 4.061333 265.6834
## ---
## 26898: 100 0.231000 238.7585
## 26899: 100 -0.710000 240.8043
## 26900: 100 -1.482000 238.6160
## 26901: 100 -1.801000 226.3945
## 26902: 100 -1.765000 233.5635
## loc4 ws_10m loc5 DSWRF_surface loc5 RH_2.m.above.ground
## 1: 2.684959 0.00 88.70000
## 2: 2.672927 0.00 88.53333
## 3: 2.660895 0.00 88.36667
## 4: 2.648862 0.00 88.20000
## 5: 2.636125 0.00 87.46667
## ---
## 26898: 4.913175 302.88 80.20000
## 26899: 4.551752 223.22 87.40000
## 26900: 4.268511 167.44 91.10000
## 26901: 3.659096 133.94 93.70000
## 26902: 3.637878 111.62 94.90000
## loc5 TCDC_low.cloud.layer loc5 TMP_2.m.above.ground loc5 wdir_10m
## 1: 2.0000000 -3.174000 336.58831
## 2: 1.3333333 -3.294667 341.74223
## 3: 0.6666667 -3.415333 346.89616
## 4: 0.0000000 -3.536000 352.05008
## 5: 0.0000000 -3.878667 349.81053
## ---
## 26898: 55.4000000 -6.329000 165.58548
## 26899: 38.6000000 -10.300000 172.49167
## 26900: 31.0000000 -12.952000 148.07612
## 26901: 26.0000000 -14.521000 100.22730
## 26902: 22.5000000 -15.455000 82.61275
## loc5 ws_10m loc6 DSWRF_surface loc6 RH_2.m.above.ground
## 1: 0.7788009 0.00 94.20000
## 2: 1.0005702 0.00 94.10000
## 3: 1.2223395 0.00 94.00000
## 4: 1.4441088 0.00 93.90000
## 5: 1.5497829 0.00 93.63333
## ---
## 26898: 2.1663980 196.60 89.50000
## 26899: 1.8864951 141.86 94.90000
## 26900: 1.2405518 106.40 96.40000
## 26901: 1.3722332 85.12 96.50000
## 26902: 1.9393671 70.94 96.70000
## loc6 TCDC_low.cloud.layer loc6 TMP_2.m.above.ground loc6 wdir_10m
## 1: 11.00000 -6.874000 93.47636
## 2: 12.00000 -7.021333 90.29830
## 3: 13.00000 -7.168667 87.12025
## 4: 14.00000 -7.316000 83.94219
## 5: 10.66667 -7.545333 93.91665
## ---
## 26898: 97.50000 -7.649000 62.27623
## 26899: 93.10000 -10.020000 45.36578
## 26900: 93.10000 -10.882000 36.68694
## 26901: 92.70000 -11.191000 19.84014
## 26902: 93.30000 -11.685000 357.44609
## loc6 ws_10m loc7 DSWRF_surface loc7 RH_2.m.above.ground
## 1: 0.9122336 0.00 76.20000
## 2: 0.8931703 0.00 75.53333
## 3: 0.8741070 0.00 74.86667
## 4: 0.8550436 0.00 74.20000
## 5: 0.7646836 0.00 73.43333
## ---
## 26898: 2.9702756 235.86 59.30000
## 26899: 1.8641058 174.56 61.00000
## 26900: 1.0812659 131.02 61.10000
## 26901: 0.7084035 104.80 59.00000
## 26902: 0.5999499 87.34 58.30000
## loc7 TCDC_low.cloud.layer loc7 TMP_2.m.above.ground loc7 wdir_10m
## 1: 83.00000 6.886000 288.4040
## 2: 76.33333 6.808667 296.8295
## 3: 69.66667 6.731333 305.2550
## 4: 63.00000 6.654000 313.6806
## 5: 57.66667 6.551333 308.5280
## ---
## 26898: 43.80000 1.081000 180.9141
## 26899: 32.70000 0.560000 181.5385
## 26900: 37.30000 0.198000 182.9638
## 26901: 49.10000 0.289000 187.1146
## 26902: 57.40000 0.275000 190.2860
## loc7 ws_10m
## 1: 3.150579
## 2: 3.118881
## 3: 3.087182
## 4: 3.055483
## 5: 2.920705
## ---
## 26898: 5.058844
## 26899: 4.972112
## 26900: 4.719232
## 26901: 4.840923
## 26902: 4.685959names(trainw2) <- gsub(" ", "_", names(trainw2))
trainw2## datetime loc1_DSWRF_surface loc1_RH_2.m.above.ground
## 1: 2018-12-31 00:00:00 0.00 83.30000
## 2: 2018-12-31 01:00:00 0.00 82.00000
## 3: 2018-12-31 02:00:00 0.00 80.70000
## 4: 2018-12-31 03:00:00 0.00 79.40000
## 5: 2018-12-31 04:00:00 0.00 78.83333
## ---
## 26898: 2022-01-24 17:00:00 348.20 75.60000
## 26899: 2022-01-24 18:00:00 255.46 79.00000
## 26900: 2022-01-24 19:00:00 191.60 78.00000
## 26901: 2022-01-24 20:00:00 153.28 80.00000
## 26902: 2022-01-24 21:00:00 127.74 79.90000
## loc1_TCDC_low.cloud.layer loc1_TMP_2.m.above.ground loc1_wdir_10m
## 1: 0.0 -8.284000 219.0959
## 2: 1.0 -8.104667 217.8237
## 3: 2.0 -7.925333 216.5515
## 4: 3.0 -7.746000 215.2793
## 5: 5.0 -7.995333 216.4846
## ---
## 26898: 3.4 -8.619000 232.9766
## 26899: 2.4 -13.750000 258.6988
## 26900: 1.8 -15.312000 261.6345
## 26901: 1.5 -15.771000 262.4866
## 26902: 1.2 -15.815000 264.9919
## loc1_ws_10m loc2_DSWRF_surface loc2_RH_2.m.above.ground
## 1: 2.790017 0.00 76.1
## 2: 2.725563 0.00 74.6
## 3: 2.661109 0.00 73.1
## 4: 2.596655 0.00 71.6
## 5: 2.551815 0.00 71.3
## ---
## 26898: 1.541491 278.72 69.0
## 26899: 2.451014 197.54 65.9
## 26900: 2.975670 148.14 67.6
## 26901: 3.469357 118.52 73.1
## 26902: 3.329440 98.78 76.8
## loc2_TCDC_low.cloud.layer loc2_TMP_2.m.above.ground loc2_wdir_10m
## 1: 66.00000 4.396000 179.5000
## 2: 59.66667 3.972000 182.0823
## 3: 53.33333 3.548000 184.6647
## 4: 47.00000 3.124000 187.2471
## 5: 31.66667 3.111333 186.7205
## ---
## 26898: 0.00000 3.611000 163.0320
## 26899: 0.00000 1.030000 168.9420
## 26900: 0.00000 0.198000 169.6947
## 26901: 0.00000 -0.651000 168.4363
## 26902: 0.00000 -1.405000 181.3601
## loc2_ws_10m loc3_DSWRF_surface loc3_RH_2.m.above.ground
## 1: 2.3553997 0.00 93.20000
## 2: 2.3094280 0.00 92.93333
## 3: 2.2634562 0.00 92.66667
## 4: 2.2174845 0.00 92.40000
## 5: 2.2073096 0.00 92.06667
## ---
## 26898: 0.3744988 308.36 81.10000
## 26899: 0.3467570 222.70 85.70000
## 26900: 1.1514952 167.02 86.10000
## 26901: 2.1472339 133.62 85.80000
## 26902: 2.8114421 111.34 84.90000
## loc3_TCDC_low.cloud.layer loc3_TMP_2.m.above.ground loc3_wdir_10m
## 1: 95.0 -4.684000 231.8005
## 2: 89.0 -4.884667 245.0924
## 3: 83.0 -5.085333 258.3843
## 4: 77.0 -5.286000 271.6762
## 5: 54.0 -5.995333 217.6031
## ---
## 26898: 79.6 -8.829000 188.7512
## 26899: 77.2 -12.350000 148.3612
## 26900: 74.2 -13.982000 127.2406
## 26901: 72.7 -14.521000 120.5353
## 26902: 73.8 -14.465000 116.5458
## loc3_ws_10m loc4_DSWRF_surface loc4_RH_2.m.above.ground
## 1: 0.8009606 0.00 75.20000
## 2: 0.6506007 0.00 76.03333
## 3: 0.5002407 0.00 76.86667
## 4: 0.3498807 0.00 77.70000
## 5: 0.3796237 0.00 77.60000
## ---
## 26898: 2.4365654 190.82 63.50000
## 26899: 2.6431705 144.14 72.80000
## 26900: 2.8966046 108.96 77.00000
## 26901: 2.8217623 87.18 76.60000
## 26902: 2.6194151 72.64 74.70000
## loc4_TCDC_low.cloud.layer loc4_TMP_2.m.above.ground loc4_wdir_10m
## 1: 0 5.026000 253.8836
## 2: 0 4.732000 257.6694
## 3: 0 4.438000 261.4552
## 4: 0 4.144000 265.2409
## 5: 0 4.061333 265.6834
## ---
## 26898: 100 0.231000 238.7585
## 26899: 100 -0.710000 240.8043
## 26900: 100 -1.482000 238.6160
## 26901: 100 -1.801000 226.3945
## 26902: 100 -1.765000 233.5635
## loc4_ws_10m loc5_DSWRF_surface loc5_RH_2.m.above.ground
## 1: 2.684959 0.00 88.70000
## 2: 2.672927 0.00 88.53333
## 3: 2.660895 0.00 88.36667
## 4: 2.648862 0.00 88.20000
## 5: 2.636125 0.00 87.46667
## ---
## 26898: 4.913175 302.88 80.20000
## 26899: 4.551752 223.22 87.40000
## 26900: 4.268511 167.44 91.10000
## 26901: 3.659096 133.94 93.70000
## 26902: 3.637878 111.62 94.90000
## loc5_TCDC_low.cloud.layer loc5_TMP_2.m.above.ground loc5_wdir_10m
## 1: 2.0000000 -3.174000 336.58831
## 2: 1.3333333 -3.294667 341.74223
## 3: 0.6666667 -3.415333 346.89616
## 4: 0.0000000 -3.536000 352.05008
## 5: 0.0000000 -3.878667 349.81053
## ---
## 26898: 55.4000000 -6.329000 165.58548
## 26899: 38.6000000 -10.300000 172.49167
## 26900: 31.0000000 -12.952000 148.07612
## 26901: 26.0000000 -14.521000 100.22730
## 26902: 22.5000000 -15.455000 82.61275
## loc5_ws_10m loc6_DSWRF_surface loc6_RH_2.m.above.ground
## 1: 0.7788009 0.00 94.20000
## 2: 1.0005702 0.00 94.10000
## 3: 1.2223395 0.00 94.00000
## 4: 1.4441088 0.00 93.90000
## 5: 1.5497829 0.00 93.63333
## ---
## 26898: 2.1663980 196.60 89.50000
## 26899: 1.8864951 141.86 94.90000
## 26900: 1.2405518 106.40 96.40000
## 26901: 1.3722332 85.12 96.50000
## 26902: 1.9393671 70.94 96.70000
## loc6_TCDC_low.cloud.layer loc6_TMP_2.m.above.ground loc6_wdir_10m
## 1: 11.00000 -6.874000 93.47636
## 2: 12.00000 -7.021333 90.29830
## 3: 13.00000 -7.168667 87.12025
## 4: 14.00000 -7.316000 83.94219
## 5: 10.66667 -7.545333 93.91665
## ---
## 26898: 97.50000 -7.649000 62.27623
## 26899: 93.10000 -10.020000 45.36578
## 26900: 93.10000 -10.882000 36.68694
## 26901: 92.70000 -11.191000 19.84014
## 26902: 93.30000 -11.685000 357.44609
## loc6_ws_10m loc7_DSWRF_surface loc7_RH_2.m.above.ground
## 1: 0.9122336 0.00 76.20000
## 2: 0.8931703 0.00 75.53333
## 3: 0.8741070 0.00 74.86667
## 4: 0.8550436 0.00 74.20000
## 5: 0.7646836 0.00 73.43333
## ---
## 26898: 2.9702756 235.86 59.30000
## 26899: 1.8641058 174.56 61.00000
## 26900: 1.0812659 131.02 61.10000
## 26901: 0.7084035 104.80 59.00000
## 26902: 0.5999499 87.34 58.30000
## loc7_TCDC_low.cloud.layer loc7_TMP_2.m.above.ground loc7_wdir_10m
## 1: 83.00000 6.886000 288.4040
## 2: 76.33333 6.808667 296.8295
## 3: 69.66667 6.731333 305.2550
## 4: 63.00000 6.654000 313.6806
## 5: 57.66667 6.551333 308.5280
## ---
## 26898: 43.80000 1.081000 180.9141
## 26899: 32.70000 0.560000 181.5385
## 26900: 37.30000 0.198000 182.9638
## 26901: 49.10000 0.289000 187.1146
## 26902: 57.40000 0.275000 190.2860
## loc7_ws_10m
## 1: 3.150579
## 2: 3.118881
## 3: 3.087182
## 4: 3.055483
## 5: 2.920705
## ---
## 26898: 5.058844
## 26899: 4.972112
## 26900: 4.719232
## 26901: 4.840923
## 26902: 4.685959all = merge(c2,trainw2, by="datetime")
all## datetime net system_direction loc1_DSWRF_surface
## 1: 2019-01-01 00:00:00 -1934.795 Negative 0.00
## 2: 2019-01-01 01:00:00 -750.866 Negative 0.00
## 3: 2019-01-01 02:00:00 -982.767 Negative 0.00
## 4: 2019-01-01 03:00:00 -764.620 Negative 0.00
## 5: 2019-01-01 04:00:00 -1233.289 Negative 0.00
## ---
## 26804: 2022-01-21 19:00:00 0.000 55.38
## 26805: 2022-01-21 20:00:00 0.000 44.30
## 26806: 2022-01-21 21:00:00 0.000 36.92
## 26807: 2022-01-21 22:00:00 0.000 0.00
## 26808: 2022-01-21 23:00:00 0.000 0.00
## loc1_RH_2.m.above.ground loc1_TCDC_low.cloud.layer
## 1: 84.20000 0.00000
## 2: 84.43333 3.00000
## 3: 84.66667 6.00000
## 4: 84.90000 9.00000
## 5: 87.26667 29.66667
## ---
## 26804: 97.90000 99.90000
## 26805: 98.00000 99.90000
## 26806: 98.70000 99.90000
## 26807: 98.80000 100.00000
## 26808: 98.90000 100.00000
## loc1_TMP_2.m.above.ground loc1_wdir_10m loc1_ws_10m loc2_DSWRF_surface
## 1: -4.081 244.986564 2.302671 0.00
## 2: -3.707 245.382871 2.443606 0.00
## 3: -3.333 245.779178 2.584541 0.00
## 4: -2.959 246.175485 2.725476 0.00
## 5: -2.591 240.651199 2.755832 0.00
## ---
## 26804: -1.623 2.212407 1.948613 40.46
## 26805: -1.285 357.806092 1.767906 32.36
## 26806: -0.937 354.741987 2.657864 26.98
## 26807: -0.641 7.030764 3.098056 0.00
## 26808: -0.540 27.723707 3.520917 0.00
## loc2_RH_2.m.above.ground loc2_TCDC_low.cloud.layer
## 1: 59.80000 0.0
## 2: 56.93333 0.0
## 3: 54.06667 0.0
## 4: 51.20000 0.0
## 5: 51.56667 0.0
## ---
## 26804: 66.80000 99.9
## 26805: 68.60000 99.9
## 26806: 68.50000 99.9
## 26807: 73.90000 100.0
## 26808: 79.80000 100.0
## loc2_TMP_2.m.above.ground loc2_wdir_10m loc2_ws_10m loc3_DSWRF_surface
## 1: 3.369000 192.5959 3.057220 0.00
## 2: 3.333000 192.2694 3.202298 0.00
## 3: 3.297000 191.9429 3.347376 0.00
## 4: 3.261000 191.6164 3.492454 0.00
## 5: 3.365667 191.7771 3.608525 0.00
## ---
## 26804: 4.347000 208.0302 3.265933 163.96
## 26805: 4.295000 203.6774 3.355885 131.16
## 26806: 4.533000 206.7301 3.698566 109.30
## 26807: 4.649000 211.2813 4.276995 0.00
## 26808: 4.530000 212.4927 3.987479 0.00
## loc3_RH_2.m.above.ground loc3_TCDC_low.cloud.layer
## 1: 84.40000 1.0000000
## 2: 85.06667 0.6666667
## 3: 85.73333 0.3333333
## 4: 86.40000 0.0000000
## 5: 87.83333 1.3333333
## ---
## 26804: 72.40000 0.0000000
## 26805: 71.90000 0.0000000
## 26806: 72.70000 0.0000000
## 26807: 74.60000 5.0000000
## 26808: 77.70000 4.8000000
## loc3_TMP_2.m.above.ground loc3_wdir_10m loc3_ws_10m loc4_DSWRF_surface
## 1: -6.211000 351.73126 1.784915 0.00
## 2: -5.950333 350.11411 1.928927 0.00
## 3: -5.689667 348.49697 2.072939 0.00
## 4: -5.429000 346.87983 2.216951 0.00
## 5: -5.094333 337.72332 1.997965 0.00
## ---
## 26804: -2.723000 43.76432 3.014659 30.56
## 26805: -3.055000 45.90085 2.955315 24.46
## 26806: -3.107000 42.20943 2.763192 20.36
## 26807: -3.091000 39.53944 2.716282 0.00
## 26808: -2.910000 51.42887 2.849656 0.00
## loc4_RH_2.m.above.ground loc4_TCDC_low.cloud.layer
## 1: 52.20000 0.00000
## 2: 53.90000 9.00000
## 3: 55.60000 18.00000
## 4: 57.30000 27.00000
## 5: 59.76667 51.33333
## ---
## 26804: 88.20000 98.10000
## 26805: 89.20000 98.50000
## 26806: 89.20000 98.70000
## 26807: 89.30000 100.00000
## 26808: 90.10000 100.00000
## loc4_TMP_2.m.above.ground loc4_wdir_10m loc4_ws_10m loc5_DSWRF_surface
## 1: 7.309000 248.9115 5.151732 0.00
## 2: 7.203000 248.8854 5.512422 0.00
## 3: 7.097000 248.8594 5.873112 0.00
## 4: 6.991000 248.8333 6.233802 0.00
## 5: 6.902333 248.9605 6.574972 0.00
## ---
## 26804: 9.287000 352.6565 4.262119 116.34
## 26805: 9.265000 352.4482 4.243415 93.08
## 26806: 9.233000 350.1066 4.036709 77.56
## 26807: 9.189000 346.2419 3.577402 0.00
## 26808: 9.140000 339.1743 3.634154 0.00
## loc5_RH_2.m.above.ground loc5_TCDC_low.cloud.layer
## 1: 88.30000 0.0000000
## 2: 87.66667 0.0000000
## 3: 87.03333 0.0000000
## 4: 86.40000 0.0000000
## 5: 86.66667 0.3333333
## ---
## 26804: 84.30000 69.8000000
## 26805: 81.30000 69.5000000
## 26806: 86.40000 70.2000000
## 26807: 89.50000 100.0000000
## 26808: 94.00000 99.1000000
## loc5_TMP_2.m.above.ground loc5_wdir_10m loc5_ws_10m loc6_DSWRF_surface
## 1: -4.191000 297.65206 3.913716 0.00
## 2: -4.070333 295.48408 4.243639 0.00
## 3: -3.949667 293.31610 4.573562 0.00
## 4: -3.829000 291.14811 4.903485 0.00
## 5: -3.864333 290.61532 4.996934 0.00
## ---
## 26804: 1.997000 19.05118 4.366314 114.36
## 26805: 2.185000 29.69346 5.556206 91.48
## 26806: 1.463000 47.68443 4.525578 76.24
## 26807: 1.679000 63.83293 3.843080 0.00
## 26808: 1.490000 56.89650 3.948929 0.00
## loc6_RH_2.m.above.ground loc6_TCDC_low.cloud.layer
## 1: 92.10000 28.00000
## 2: 92.50000 25.66667
## 3: 92.90000 23.33333
## 4: 93.30000 21.00000
## 5: 94.03333 21.00000
## ---
## 26804: 73.20000 74.30000
## 26805: 72.20000 79.40000
## 26806: 74.20000 82.60000
## 26807: 89.20000 100.00000
## 26808: 82.50000 100.00000
## loc6_TMP_2.m.above.ground loc6_wdir_10m loc6_ws_10m loc7_DSWRF_surface
## 1: -7.041000 280.4415 1.745607 0.00
## 2: -7.200333 274.3150 1.962468 0.00
## 3: -7.359667 268.1885 2.179329 0.00
## 4: -7.519000 262.0619 2.396190 0.00
## 5: -7.527667 260.3688 2.522906 0.00
## ---
## 26804: -1.953000 324.6584 2.117349 7.32
## 26805: -1.895000 312.2008 2.332197 5.86
## 26806: -1.907000 331.4124 2.160014 4.88
## 26807: -2.371000 300.8462 2.388677 0.00
## 26808: -1.210000 345.0344 2.563677 0.00
## loc7_RH_2.m.above.ground loc7_TCDC_low.cloud.layer
## 1: 65.80000 0
## 2: 66.40000 0
## 3: 67.00000 0
## 4: 67.60000 0
## 5: 67.83333 0
## ---
## 26804: 94.90000 100
## 26805: 94.80000 100
## 26806: 95.00000 100
## 26807: 95.20000 100
## 26808: 95.20000 100
## loc7_TMP_2.m.above.ground loc7_wdir_10m loc7_ws_10m
## 1: 7.239000 250.5319 5.861836
## 2: 7.153000 250.9196 6.254371
## 3: 7.067000 251.3073 6.646906
## 4: 6.981000 251.6950 7.039441
## 5: 6.962333 252.8526 7.291198
## ---
## 26804: 3.577000 192.1873 6.417472
## 26805: 2.815000 181.7018 6.656326
## 26806: 2.743000 185.1566 6.047797
## 26807: 2.399000 184.1567 6.081236
## 26808: 2.180000 182.7795 6.641093ggplot(all ,aes(x=datetime,y=net)) + geom_line()
all[,trnd:=1:.N]
all[,w_day:=as.character(wday(datetime,label=T))]
all[,mon:=as.character(month(datetime,label=T))]
all[,hour:=as.factor(hour(datetime))]
all[,Date:=as.Date(datetime)]
all## datetime net system_direction loc1_DSWRF_surface
## 1: 2019-01-01 00:00:00 -1934.795 Negative 0.00
## 2: 2019-01-01 01:00:00 -750.866 Negative 0.00
## 3: 2019-01-01 02:00:00 -982.767 Negative 0.00
## 4: 2019-01-01 03:00:00 -764.620 Negative 0.00
## 5: 2019-01-01 04:00:00 -1233.289 Negative 0.00
## ---
## 26804: 2022-01-21 19:00:00 0.000 55.38
## 26805: 2022-01-21 20:00:00 0.000 44.30
## 26806: 2022-01-21 21:00:00 0.000 36.92
## 26807: 2022-01-21 22:00:00 0.000 0.00
## 26808: 2022-01-21 23:00:00 0.000 0.00
## loc1_RH_2.m.above.ground loc1_TCDC_low.cloud.layer
## 1: 84.20000 0.00000
## 2: 84.43333 3.00000
## 3: 84.66667 6.00000
## 4: 84.90000 9.00000
## 5: 87.26667 29.66667
## ---
## 26804: 97.90000 99.90000
## 26805: 98.00000 99.90000
## 26806: 98.70000 99.90000
## 26807: 98.80000 100.00000
## 26808: 98.90000 100.00000
## loc1_TMP_2.m.above.ground loc1_wdir_10m loc1_ws_10m loc2_DSWRF_surface
## 1: -4.081 244.986564 2.302671 0.00
## 2: -3.707 245.382871 2.443606 0.00
## 3: -3.333 245.779178 2.584541 0.00
## 4: -2.959 246.175485 2.725476 0.00
## 5: -2.591 240.651199 2.755832 0.00
## ---
## 26804: -1.623 2.212407 1.948613 40.46
## 26805: -1.285 357.806092 1.767906 32.36
## 26806: -0.937 354.741987 2.657864 26.98
## 26807: -0.641 7.030764 3.098056 0.00
## 26808: -0.540 27.723707 3.520917 0.00
## loc2_RH_2.m.above.ground loc2_TCDC_low.cloud.layer
## 1: 59.80000 0.0
## 2: 56.93333 0.0
## 3: 54.06667 0.0
## 4: 51.20000 0.0
## 5: 51.56667 0.0
## ---
## 26804: 66.80000 99.9
## 26805: 68.60000 99.9
## 26806: 68.50000 99.9
## 26807: 73.90000 100.0
## 26808: 79.80000 100.0
## loc2_TMP_2.m.above.ground loc2_wdir_10m loc2_ws_10m loc3_DSWRF_surface
## 1: 3.369000 192.5959 3.057220 0.00
## 2: 3.333000 192.2694 3.202298 0.00
## 3: 3.297000 191.9429 3.347376 0.00
## 4: 3.261000 191.6164 3.492454 0.00
## 5: 3.365667 191.7771 3.608525 0.00
## ---
## 26804: 4.347000 208.0302 3.265933 163.96
## 26805: 4.295000 203.6774 3.355885 131.16
## 26806: 4.533000 206.7301 3.698566 109.30
## 26807: 4.649000 211.2813 4.276995 0.00
## 26808: 4.530000 212.4927 3.987479 0.00
## loc3_RH_2.m.above.ground loc3_TCDC_low.cloud.layer
## 1: 84.40000 1.0000000
## 2: 85.06667 0.6666667
## 3: 85.73333 0.3333333
## 4: 86.40000 0.0000000
## 5: 87.83333 1.3333333
## ---
## 26804: 72.40000 0.0000000
## 26805: 71.90000 0.0000000
## 26806: 72.70000 0.0000000
## 26807: 74.60000 5.0000000
## 26808: 77.70000 4.8000000
## loc3_TMP_2.m.above.ground loc3_wdir_10m loc3_ws_10m loc4_DSWRF_surface
## 1: -6.211000 351.73126 1.784915 0.00
## 2: -5.950333 350.11411 1.928927 0.00
## 3: -5.689667 348.49697 2.072939 0.00
## 4: -5.429000 346.87983 2.216951 0.00
## 5: -5.094333 337.72332 1.997965 0.00
## ---
## 26804: -2.723000 43.76432 3.014659 30.56
## 26805: -3.055000 45.90085 2.955315 24.46
## 26806: -3.107000 42.20943 2.763192 20.36
## 26807: -3.091000 39.53944 2.716282 0.00
## 26808: -2.910000 51.42887 2.849656 0.00
## loc4_RH_2.m.above.ground loc4_TCDC_low.cloud.layer
## 1: 52.20000 0.00000
## 2: 53.90000 9.00000
## 3: 55.60000 18.00000
## 4: 57.30000 27.00000
## 5: 59.76667 51.33333
## ---
## 26804: 88.20000 98.10000
## 26805: 89.20000 98.50000
## 26806: 89.20000 98.70000
## 26807: 89.30000 100.00000
## 26808: 90.10000 100.00000
## loc4_TMP_2.m.above.ground loc4_wdir_10m loc4_ws_10m loc5_DSWRF_surface
## 1: 7.309000 248.9115 5.151732 0.00
## 2: 7.203000 248.8854 5.512422 0.00
## 3: 7.097000 248.8594 5.873112 0.00
## 4: 6.991000 248.8333 6.233802 0.00
## 5: 6.902333 248.9605 6.574972 0.00
## ---
## 26804: 9.287000 352.6565 4.262119 116.34
## 26805: 9.265000 352.4482 4.243415 93.08
## 26806: 9.233000 350.1066 4.036709 77.56
## 26807: 9.189000 346.2419 3.577402 0.00
## 26808: 9.140000 339.1743 3.634154 0.00
## loc5_RH_2.m.above.ground loc5_TCDC_low.cloud.layer
## 1: 88.30000 0.0000000
## 2: 87.66667 0.0000000
## 3: 87.03333 0.0000000
## 4: 86.40000 0.0000000
## 5: 86.66667 0.3333333
## ---
## 26804: 84.30000 69.8000000
## 26805: 81.30000 69.5000000
## 26806: 86.40000 70.2000000
## 26807: 89.50000 100.0000000
## 26808: 94.00000 99.1000000
## loc5_TMP_2.m.above.ground loc5_wdir_10m loc5_ws_10m loc6_DSWRF_surface
## 1: -4.191000 297.65206 3.913716 0.00
## 2: -4.070333 295.48408 4.243639 0.00
## 3: -3.949667 293.31610 4.573562 0.00
## 4: -3.829000 291.14811 4.903485 0.00
## 5: -3.864333 290.61532 4.996934 0.00
## ---
## 26804: 1.997000 19.05118 4.366314 114.36
## 26805: 2.185000 29.69346 5.556206 91.48
## 26806: 1.463000 47.68443 4.525578 76.24
## 26807: 1.679000 63.83293 3.843080 0.00
## 26808: 1.490000 56.89650 3.948929 0.00
## loc6_RH_2.m.above.ground loc6_TCDC_low.cloud.layer
## 1: 92.10000 28.00000
## 2: 92.50000 25.66667
## 3: 92.90000 23.33333
## 4: 93.30000 21.00000
## 5: 94.03333 21.00000
## ---
## 26804: 73.20000 74.30000
## 26805: 72.20000 79.40000
## 26806: 74.20000 82.60000
## 26807: 89.20000 100.00000
## 26808: 82.50000 100.00000
## loc6_TMP_2.m.above.ground loc6_wdir_10m loc6_ws_10m loc7_DSWRF_surface
## 1: -7.041000 280.4415 1.745607 0.00
## 2: -7.200333 274.3150 1.962468 0.00
## 3: -7.359667 268.1885 2.179329 0.00
## 4: -7.519000 262.0619 2.396190 0.00
## 5: -7.527667 260.3688 2.522906 0.00
## ---
## 26804: -1.953000 324.6584 2.117349 7.32
## 26805: -1.895000 312.2008 2.332197 5.86
## 26806: -1.907000 331.4124 2.160014 4.88
## 26807: -2.371000 300.8462 2.388677 0.00
## 26808: -1.210000 345.0344 2.563677 0.00
## loc7_RH_2.m.above.ground loc7_TCDC_low.cloud.layer
## 1: 65.80000 0
## 2: 66.40000 0
## 3: 67.00000 0
## 4: 67.60000 0
## 5: 67.83333 0
## ---
## 26804: 94.90000 100
## 26805: 94.80000 100
## 26806: 95.00000 100
## 26807: 95.20000 100
## 26808: 95.20000 100
## loc7_TMP_2.m.above.ground loc7_wdir_10m loc7_ws_10m trnd w_day mon hour
## 1: 7.239000 250.5319 5.861836 1 Sal Oca 0
## 2: 7.153000 250.9196 6.254371 2 Sal Oca 1
## 3: 7.067000 251.3073 6.646906 3 Sal Oca 2
## 4: 6.981000 251.6950 7.039441 4 Sal Oca 3
## 5: 6.962333 252.8526 7.291198 5 Sal Oca 4
## ---
## 26804: 3.577000 192.1873 6.417472 26804 Cum Oca 19
## 26805: 2.815000 181.7018 6.656326 26805 Cum Oca 20
## 26806: 2.743000 185.1566 6.047797 26806 Cum Oca 21
## 26807: 2.399000 184.1567 6.081236 26807 Cum Oca 22
## 26808: 2.180000 182.7795 6.641093 26808 Cum Oca 23
## Date
## 1: 2019-01-01
## 2: 2019-01-01
## 3: 2019-01-01
## 4: 2019-01-01
## 5: 2019-01-01
## ---
## 26804: 2022-01-21
## 26805: 2022-01-21
## 26806: 2022-01-21
## 26807: 2022-01-21
## 26808: 2022-01-21First linear regression uses all parameters and it’s our baseline
lm1=lm(net~trnd+w_day+mon+hour+
loc1_DSWRF_surface+loc1_RH_2.m.above.ground+loc1_TCDC_low.cloud.layer+loc1_TMP_2.m.above.ground+loc1_wdir_10m+loc1_ws_10m+
loc2_DSWRF_surface+loc2_RH_2.m.above.ground+loc2_TCDC_low.cloud.layer+loc2_TMP_2.m.above.ground+loc2_wdir_10m+loc1_ws_10m+
loc3_DSWRF_surface+loc3_RH_2.m.above.ground+loc3_TCDC_low.cloud.layer+loc3_TMP_2.m.above.ground+loc3_wdir_10m+loc1_ws_10m+
loc4_DSWRF_surface+loc4_RH_2.m.above.ground+loc4_TCDC_low.cloud.layer+loc4_TMP_2.m.above.ground+loc4_wdir_10m+loc1_ws_10m+
loc5_DSWRF_surface+loc5_RH_2.m.above.ground+loc5_TCDC_low.cloud.layer+loc5_TMP_2.m.above.ground+loc5_wdir_10m+loc1_ws_10m+
loc6_DSWRF_surface+loc6_RH_2.m.above.ground+loc6_TCDC_low.cloud.layer+loc6_TMP_2.m.above.ground+loc6_wdir_10m+loc1_ws_10m+
loc7_DSWRF_surface+loc7_RH_2.m.above.ground+loc7_TCDC_low.cloud.layer+loc7_TMP_2.m.above.ground+loc7_wdir_10m+loc1_ws_10m
,all)
summary(lm1)##
## Call:
## lm(formula = net ~ trnd + w_day + mon + hour + loc1_DSWRF_surface +
## loc1_RH_2.m.above.ground + loc1_TCDC_low.cloud.layer + loc1_TMP_2.m.above.ground +
## loc1_wdir_10m + loc1_ws_10m + loc2_DSWRF_surface + loc2_RH_2.m.above.ground +
## loc2_TCDC_low.cloud.layer + loc2_TMP_2.m.above.ground + loc2_wdir_10m +
## loc1_ws_10m + loc3_DSWRF_surface + loc3_RH_2.m.above.ground +
## loc3_TCDC_low.cloud.layer + loc3_TMP_2.m.above.ground + loc3_wdir_10m +
## loc1_ws_10m + loc4_DSWRF_surface + loc4_RH_2.m.above.ground +
## loc4_TCDC_low.cloud.layer + loc4_TMP_2.m.above.ground + loc4_wdir_10m +
## loc1_ws_10m + loc5_DSWRF_surface + loc5_RH_2.m.above.ground +
## loc5_TCDC_low.cloud.layer + loc5_TMP_2.m.above.ground + loc5_wdir_10m +
## loc1_ws_10m + loc6_DSWRF_surface + loc6_RH_2.m.above.ground +
## loc6_TCDC_low.cloud.layer + loc6_TMP_2.m.above.ground + loc6_wdir_10m +
## loc1_ws_10m + loc7_DSWRF_surface + loc7_RH_2.m.above.ground +
## loc7_TCDC_low.cloud.layer + loc7_TMP_2.m.above.ground + loc7_wdir_10m +
## loc1_ws_10m, data = all)
##
## Residuals:
## Min 1Q Median 3Q Max
## -5672.1 -561.8 -21.7 515.9 8418.0
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 3.061e+02 1.330e+02 2.301 0.021383 *
## trnd 2.642e-02 8.691e-04 30.404 < 2e-16 ***
## w_dayCum 1.138e+02 2.231e+01 5.100 3.42e-07 ***
## w_dayÇar 1.367e+02 2.235e+01 6.116 9.75e-10 ***
## w_dayPaz -2.219e+02 2.233e+01 -9.935 < 2e-16 ***
## w_dayPer 1.857e+02 2.234e+01 8.313 < 2e-16 ***
## w_dayPts 1.031e+02 2.235e+01 4.615 3.94e-06 ***
## w_daySal 1.462e+02 2.238e+01 6.532 6.62e-11 ***
## monAra 1.578e+02 5.644e+01 2.797 0.005169 **
## monEki -1.548e+02 3.776e+01 -4.100 4.15e-05 ***
## monEyl -3.139e+02 3.138e+01 -10.004 < 2e-16 ***
## monHaz -4.143e+01 3.255e+01 -1.273 0.203068
## monKas -3.717e+01 4.716e+01 -0.788 0.430591
## monMar 8.045e+01 5.689e+01 1.414 0.157381
## monMay -1.587e+02 3.874e+01 -4.097 4.20e-05 ***
## monNis -9.667e+01 5.095e+01 -1.897 0.057811 .
## monOca 2.050e+02 6.087e+01 3.368 0.000757 ***
## monŞub 2.378e+02 6.127e+01 3.881 0.000104 ***
## monTem 6.148e+01 2.962e+01 2.076 0.037925 *
## hour1 1.785e+02 4.115e+01 4.338 1.44e-05 ***
## hour2 1.338e+02 4.125e+01 3.243 0.001184 **
## hour3 9.590e+01 4.138e+01 2.318 0.020468 *
## hour4 4.285e+01 4.158e+01 1.031 0.302735
## hour5 3.605e+01 4.176e+01 0.863 0.387913
## hour6 -5.076e+01 4.191e+01 -1.211 0.225812
## hour7 -2.432e+01 4.193e+01 -0.580 0.561870
## hour8 9.537e+01 4.200e+01 2.271 0.023164 *
## hour9 2.774e+02 4.268e+01 6.499 8.22e-11 ***
## hour10 3.014e+02 5.020e+01 6.005 1.94e-09 ***
## hour11 3.660e+02 5.265e+01 6.951 3.71e-12 ***
## hour12 1.156e+02 5.490e+01 2.105 0.035289 *
## hour13 3.037e+00 5.651e+01 0.054 0.957137
## hour14 2.335e+02 5.738e+01 4.070 4.71e-05 ***
## hour15 2.024e+02 5.745e+01 3.523 0.000428 ***
## hour16 6.460e+01 5.420e+01 1.192 0.233274
## hour17 1.578e+02 5.154e+01 3.061 0.002208 **
## hour18 1.723e+02 4.869e+01 3.538 0.000403 ***
## hour19 1.999e+02 4.637e+01 4.311 1.63e-05 ***
## hour20 2.415e+02 4.482e+01 5.389 7.15e-08 ***
## hour21 3.197e+02 4.369e+01 7.317 2.60e-13 ***
## hour22 2.126e+02 4.146e+01 5.129 2.94e-07 ***
## hour23 1.854e+02 4.117e+01 4.503 6.73e-06 ***
## loc1_DSWRF_surface 5.437e-01 1.155e-01 4.706 2.53e-06 ***
## loc1_RH_2.m.above.ground -3.633e+00 5.952e-01 -6.104 1.05e-09 ***
## loc1_TCDC_low.cloud.layer 2.962e+00 3.452e-01 8.581 < 2e-16 ***
## loc1_TMP_2.m.above.ground -3.208e+01 4.568e+00 -7.023 2.23e-12 ***
## loc1_wdir_10m 1.496e-01 7.659e-02 1.954 0.050764 .
## loc1_ws_10m -1.665e+00 4.278e+00 -0.389 0.697177
## loc2_DSWRF_surface -7.063e-01 1.221e-01 -5.786 7.29e-09 ***
## loc2_RH_2.m.above.ground -7.273e-01 4.972e-01 -1.463 0.143503
## loc2_TCDC_low.cloud.layer 5.290e-01 2.920e-01 1.811 0.070079 .
## loc2_TMP_2.m.above.ground 8.073e-01 3.855e+00 0.209 0.834117
## loc2_wdir_10m -1.918e-01 7.083e-02 -2.708 0.006783 **
## loc3_DSWRF_surface 3.007e-01 1.194e-01 2.519 0.011769 *
## loc3_RH_2.m.above.ground 1.990e-01 8.581e-01 0.232 0.816637
## loc3_TCDC_low.cloud.layer 2.124e+00 3.265e-01 6.505 7.89e-11 ***
## loc3_TMP_2.m.above.ground 1.043e+01 5.971e+00 1.747 0.080575 .
## loc3_wdir_10m 2.447e-01 6.793e-02 3.602 0.000317 ***
## loc4_DSWRF_surface -3.900e-01 1.329e-01 -2.934 0.003353 **
## loc4_RH_2.m.above.ground -6.954e+00 8.225e-01 -8.455 < 2e-16 ***
## loc4_TCDC_low.cloud.layer -8.134e-01 3.249e-01 -2.504 0.012296 *
## loc4_TMP_2.m.above.ground -7.469e+00 4.516e+00 -1.654 0.098185 .
## loc4_wdir_10m -7.713e-01 8.016e-02 -9.622 < 2e-16 ***
## loc5_DSWRF_surface -1.719e-01 1.273e-01 -1.351 0.176841
## loc5_RH_2.m.above.ground -5.195e+00 8.985e-01 -5.782 7.48e-09 ***
## loc5_TCDC_low.cloud.layer 1.713e+00 2.917e-01 5.873 4.34e-09 ***
## loc5_TMP_2.m.above.ground 5.373e-01 6.041e+00 0.089 0.929121
## loc5_wdir_10m -1.176e-01 8.556e-02 -1.375 0.169190
## loc6_DSWRF_surface -3.349e-01 1.203e-01 -2.785 0.005355 **
## loc6_RH_2.m.above.ground -4.927e+00 8.362e-01 -5.892 3.85e-09 ***
## loc6_TCDC_low.cloud.layer 1.591e+00 3.013e-01 5.280 1.30e-07 ***
## loc6_TMP_2.m.above.ground 2.841e-01 5.839e+00 0.049 0.961188
## loc6_wdir_10m -4.353e-01 7.186e-02 -6.058 1.40e-09 ***
## loc7_DSWRF_surface 2.542e-01 1.108e-01 2.296 0.021706 *
## loc7_RH_2.m.above.ground 5.283e+00 8.485e-01 6.226 4.87e-10 ***
## loc7_TCDC_low.cloud.layer -1.679e-01 2.426e-01 -0.692 0.489042
## loc7_TMP_2.m.above.ground 4.452e+01 4.753e+00 9.368 < 2e-16 ***
## loc7_wdir_10m 6.462e-01 7.793e-02 8.292 < 2e-16 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 971.7 on 26730 degrees of freedom
## Multiple R-squared: 0.1472, Adjusted R-squared: 0.1448
## F-statistic: 59.94 on 77 and 26730 DF, p-value: < 2.2e-16tmp=copy(all)
tmp[,actual:=net]
tmp[,predicted:=predict(lm1,tmp)]
#head(tmp)
ggplot(tmp[Date=='2019-05-08'] ,aes(x=datetime)) +
geom_line(aes(y=actual,color='real')) +
geom_line(aes(y=predicted,color='predicted'))
lm2=lm(net~trnd+w_day+mon+hour,all)
summary(lm2)##
## Call:
## lm(formula = net ~ trnd + w_day + mon + hour, data = all)
##
## Residuals:
## Min 1Q Median 3Q Max
## -5492.6 -566.8 -28.1 515.0 8524.9
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) -1.964e+01 4.050e+01 -0.485 0.627711
## trnd 2.567e-02 8.178e-04 31.390 < 2e-16 ***
## w_dayCum 1.189e+02 2.270e+01 5.237 1.64e-07 ***
## w_dayÇar 1.496e+02 2.270e+01 6.591 4.46e-11 ***
## w_dayPaz -2.172e+02 2.273e+01 -9.554 < 2e-16 ***
## w_dayPer 1.939e+02 2.270e+01 8.542 < 2e-16 ***
## w_dayPts 1.049e+02 2.274e+01 4.615 3.95e-06 ***
## w_daySal 1.760e+02 2.270e+01 7.751 9.43e-15 ***
## monAra -3.697e+02 2.982e+01 -12.397 < 2e-16 ***
## monEki -3.550e+02 2.975e+01 -11.931 < 2e-16 ***
## monEyl -4.568e+02 2.998e+01 -15.237 < 2e-16 ***
## monHaz -2.484e+02 3.000e+01 -8.283 < 2e-16 ***
## monKas -3.635e+02 3.002e+01 -12.108 < 2e-16 ***
## monMar -3.812e+02 2.988e+01 -12.761 < 2e-16 ***
## monMay -4.457e+02 2.978e+01 -14.965 < 2e-16 ***
## monNis -5.232e+02 3.007e+01 -17.399 < 2e-16 ***
## monOca -4.152e+02 2.837e+01 -14.635 < 2e-16 ***
## monŞub -3.170e+02 3.063e+01 -10.352 < 2e-16 ***
## monTem -1.275e+01 2.973e+01 -0.429 0.668041
## hour1 1.531e+02 4.201e+01 3.643 0.000270 ***
## hour2 8.692e+01 4.201e+01 2.069 0.038570 *
## hour3 3.182e+01 4.201e+01 0.757 0.448900
## hour4 -2.721e+01 4.201e+01 -0.648 0.517218
## hour5 -5.142e+01 4.201e+01 -1.224 0.221019
## hour6 -1.513e+02 4.201e+01 -3.602 0.000317 ***
## hour7 -1.196e+02 4.201e+01 -2.847 0.004417 **
## hour8 5.141e+01 4.201e+01 1.224 0.221141
## hour9 3.091e+02 4.201e+01 7.358 1.93e-13 ***
## hour10 2.900e+02 4.201e+01 6.902 5.26e-12 ***
## hour11 4.235e+02 4.201e+01 10.081 < 2e-16 ***
## hour12 2.263e+02 4.201e+01 5.386 7.28e-08 ***
## hour13 1.566e+02 4.201e+01 3.726 0.000195 ***
## hour14 4.204e+02 4.201e+01 10.006 < 2e-16 ***
## hour15 4.109e+02 4.201e+01 9.780 < 2e-16 ***
## hour16 3.474e+02 4.201e+01 8.270 < 2e-16 ***
## hour17 4.369e+02 4.201e+01 10.398 < 2e-16 ***
## hour18 4.202e+02 4.201e+01 10.001 < 2e-16 ***
## hour19 3.931e+02 4.201e+01 9.357 < 2e-16 ***
## hour20 3.822e+02 4.201e+01 9.098 < 2e-16 ***
## hour21 4.084e+02 4.201e+01 9.719 < 2e-16 ***
## hour22 2.945e+02 4.201e+01 7.009 2.45e-12 ***
## hour23 2.189e+02 4.201e+01 5.209 1.91e-07 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 992.9 on 26766 degrees of freedom
## Multiple R-squared: 0.1084, Adjusted R-squared: 0.107
## F-statistic: 79.34 on 41 and 26766 DF, p-value: < 2.2e-16tmp2=copy(all)
tmp2[,actual:=net]
tmp2[,predicted:=predict(lm2,tmp2)]
#head(tmp)
ggplot(tmp2[Date=='2019-05-08'] ,aes(x=datetime)) +
geom_line(aes(y=actual,color='real')) +
geom_line(aes(y=predicted,color='predicted'))
checkresiduals(lm2$residuals)## Warning in modeldf.default(object): Could not find appropriate degrees of
## freedom for this model.
tmp2[,prevday:=shift(x=tmp2$net,n=24L,fill=mean(tmp2$net))]
tmp2[,prevweek:=shift(x=tmp2$net,n=168L,fill=mean(tmp2$net))]
tmp2## datetime net system_direction loc1_DSWRF_surface
## 1: 2019-01-01 00:00:00 -1934.795 Negative 0.00
## 2: 2019-01-01 01:00:00 -750.866 Negative 0.00
## 3: 2019-01-01 02:00:00 -982.767 Negative 0.00
## 4: 2019-01-01 03:00:00 -764.620 Negative 0.00
## 5: 2019-01-01 04:00:00 -1233.289 Negative 0.00
## ---
## 26804: 2022-01-21 19:00:00 0.000 55.38
## 26805: 2022-01-21 20:00:00 0.000 44.30
## 26806: 2022-01-21 21:00:00 0.000 36.92
## 26807: 2022-01-21 22:00:00 0.000 0.00
## 26808: 2022-01-21 23:00:00 0.000 0.00
## loc1_RH_2.m.above.ground loc1_TCDC_low.cloud.layer
## 1: 84.20000 0.00000
## 2: 84.43333 3.00000
## 3: 84.66667 6.00000
## 4: 84.90000 9.00000
## 5: 87.26667 29.66667
## ---
## 26804: 97.90000 99.90000
## 26805: 98.00000 99.90000
## 26806: 98.70000 99.90000
## 26807: 98.80000 100.00000
## 26808: 98.90000 100.00000
## loc1_TMP_2.m.above.ground loc1_wdir_10m loc1_ws_10m loc2_DSWRF_surface
## 1: -4.081 244.986564 2.302671 0.00
## 2: -3.707 245.382871 2.443606 0.00
## 3: -3.333 245.779178 2.584541 0.00
## 4: -2.959 246.175485 2.725476 0.00
## 5: -2.591 240.651199 2.755832 0.00
## ---
## 26804: -1.623 2.212407 1.948613 40.46
## 26805: -1.285 357.806092 1.767906 32.36
## 26806: -0.937 354.741987 2.657864 26.98
## 26807: -0.641 7.030764 3.098056 0.00
## 26808: -0.540 27.723707 3.520917 0.00
## loc2_RH_2.m.above.ground loc2_TCDC_low.cloud.layer
## 1: 59.80000 0.0
## 2: 56.93333 0.0
## 3: 54.06667 0.0
## 4: 51.20000 0.0
## 5: 51.56667 0.0
## ---
## 26804: 66.80000 99.9
## 26805: 68.60000 99.9
## 26806: 68.50000 99.9
## 26807: 73.90000 100.0
## 26808: 79.80000 100.0
## loc2_TMP_2.m.above.ground loc2_wdir_10m loc2_ws_10m loc3_DSWRF_surface
## 1: 3.369000 192.5959 3.057220 0.00
## 2: 3.333000 192.2694 3.202298 0.00
## 3: 3.297000 191.9429 3.347376 0.00
## 4: 3.261000 191.6164 3.492454 0.00
## 5: 3.365667 191.7771 3.608525 0.00
## ---
## 26804: 4.347000 208.0302 3.265933 163.96
## 26805: 4.295000 203.6774 3.355885 131.16
## 26806: 4.533000 206.7301 3.698566 109.30
## 26807: 4.649000 211.2813 4.276995 0.00
## 26808: 4.530000 212.4927 3.987479 0.00
## loc3_RH_2.m.above.ground loc3_TCDC_low.cloud.layer
## 1: 84.40000 1.0000000
## 2: 85.06667 0.6666667
## 3: 85.73333 0.3333333
## 4: 86.40000 0.0000000
## 5: 87.83333 1.3333333
## ---
## 26804: 72.40000 0.0000000
## 26805: 71.90000 0.0000000
## 26806: 72.70000 0.0000000
## 26807: 74.60000 5.0000000
## 26808: 77.70000 4.8000000
## loc3_TMP_2.m.above.ground loc3_wdir_10m loc3_ws_10m loc4_DSWRF_surface
## 1: -6.211000 351.73126 1.784915 0.00
## 2: -5.950333 350.11411 1.928927 0.00
## 3: -5.689667 348.49697 2.072939 0.00
## 4: -5.429000 346.87983 2.216951 0.00
## 5: -5.094333 337.72332 1.997965 0.00
## ---
## 26804: -2.723000 43.76432 3.014659 30.56
## 26805: -3.055000 45.90085 2.955315 24.46
## 26806: -3.107000 42.20943 2.763192 20.36
## 26807: -3.091000 39.53944 2.716282 0.00
## 26808: -2.910000 51.42887 2.849656 0.00
## loc4_RH_2.m.above.ground loc4_TCDC_low.cloud.layer
## 1: 52.20000 0.00000
## 2: 53.90000 9.00000
## 3: 55.60000 18.00000
## 4: 57.30000 27.00000
## 5: 59.76667 51.33333
## ---
## 26804: 88.20000 98.10000
## 26805: 89.20000 98.50000
## 26806: 89.20000 98.70000
## 26807: 89.30000 100.00000
## 26808: 90.10000 100.00000
## loc4_TMP_2.m.above.ground loc4_wdir_10m loc4_ws_10m loc5_DSWRF_surface
## 1: 7.309000 248.9115 5.151732 0.00
## 2: 7.203000 248.8854 5.512422 0.00
## 3: 7.097000 248.8594 5.873112 0.00
## 4: 6.991000 248.8333 6.233802 0.00
## 5: 6.902333 248.9605 6.574972 0.00
## ---
## 26804: 9.287000 352.6565 4.262119 116.34
## 26805: 9.265000 352.4482 4.243415 93.08
## 26806: 9.233000 350.1066 4.036709 77.56
## 26807: 9.189000 346.2419 3.577402 0.00
## 26808: 9.140000 339.1743 3.634154 0.00
## loc5_RH_2.m.above.ground loc5_TCDC_low.cloud.layer
## 1: 88.30000 0.0000000
## 2: 87.66667 0.0000000
## 3: 87.03333 0.0000000
## 4: 86.40000 0.0000000
## 5: 86.66667 0.3333333
## ---
## 26804: 84.30000 69.8000000
## 26805: 81.30000 69.5000000
## 26806: 86.40000 70.2000000
## 26807: 89.50000 100.0000000
## 26808: 94.00000 99.1000000
## loc5_TMP_2.m.above.ground loc5_wdir_10m loc5_ws_10m loc6_DSWRF_surface
## 1: -4.191000 297.65206 3.913716 0.00
## 2: -4.070333 295.48408 4.243639 0.00
## 3: -3.949667 293.31610 4.573562 0.00
## 4: -3.829000 291.14811 4.903485 0.00
## 5: -3.864333 290.61532 4.996934 0.00
## ---
## 26804: 1.997000 19.05118 4.366314 114.36
## 26805: 2.185000 29.69346 5.556206 91.48
## 26806: 1.463000 47.68443 4.525578 76.24
## 26807: 1.679000 63.83293 3.843080 0.00
## 26808: 1.490000 56.89650 3.948929 0.00
## loc6_RH_2.m.above.ground loc6_TCDC_low.cloud.layer
## 1: 92.10000 28.00000
## 2: 92.50000 25.66667
## 3: 92.90000 23.33333
## 4: 93.30000 21.00000
## 5: 94.03333 21.00000
## ---
## 26804: 73.20000 74.30000
## 26805: 72.20000 79.40000
## 26806: 74.20000 82.60000
## 26807: 89.20000 100.00000
## 26808: 82.50000 100.00000
## loc6_TMP_2.m.above.ground loc6_wdir_10m loc6_ws_10m loc7_DSWRF_surface
## 1: -7.041000 280.4415 1.745607 0.00
## 2: -7.200333 274.3150 1.962468 0.00
## 3: -7.359667 268.1885 2.179329 0.00
## 4: -7.519000 262.0619 2.396190 0.00
## 5: -7.527667 260.3688 2.522906 0.00
## ---
## 26804: -1.953000 324.6584 2.117349 7.32
## 26805: -1.895000 312.2008 2.332197 5.86
## 26806: -1.907000 331.4124 2.160014 4.88
## 26807: -2.371000 300.8462 2.388677 0.00
## 26808: -1.210000 345.0344 2.563677 0.00
## loc7_RH_2.m.above.ground loc7_TCDC_low.cloud.layer
## 1: 65.80000 0
## 2: 66.40000 0
## 3: 67.00000 0
## 4: 67.60000 0
## 5: 67.83333 0
## ---
## 26804: 94.90000 100
## 26805: 94.80000 100
## 26806: 95.00000 100
## 26807: 95.20000 100
## 26808: 95.20000 100
## loc7_TMP_2.m.above.ground loc7_wdir_10m loc7_ws_10m trnd w_day mon hour
## 1: 7.239000 250.5319 5.861836 1 Sal Oca 0
## 2: 7.153000 250.9196 6.254371 2 Sal Oca 1
## 3: 7.067000 251.3073 6.646906 3 Sal Oca 2
## 4: 6.981000 251.6950 7.039441 4 Sal Oca 3
## 5: 6.962333 252.8526 7.291198 5 Sal Oca 4
## ---
## 26804: 3.577000 192.1873 6.417472 26804 Cum Oca 19
## 26805: 2.815000 181.7018 6.656326 26805 Cum Oca 20
## 26806: 2.743000 185.1566 6.047797 26806 Cum Oca 21
## 26807: 2.399000 184.1567 6.081236 26807 Cum Oca 22
## 26808: 2.180000 182.7795 6.641093 26808 Cum Oca 23
## Date actual predicted prevday prevweek
## 1: 2019-01-01 -1934.795 -258.8107 287.9239 287.9239
## 2: 2019-01-01 -750.866 -105.7097 287.9239 287.9239
## 3: 2019-01-01 -982.767 -171.8369 287.9239 287.9239
## 4: 2019-01-01 -764.620 -226.9174 287.9239 287.9239
## 5: 2019-01-01 -1233.289 -285.9188 287.9239 287.9239
## ---
## 26804: 2022-01-21 0.000 765.3269 2043.4290 -425.3450
## 26805: 2022-01-21 0.000 754.4451 2171.4590 -339.5620
## 26806: 2022-01-21 0.000 780.6004 2981.0760 -356.4620
## 26807: 2022-01-21 0.000 666.7590 2627.5980 -522.6550
## 26808: 2022-01-21 0.000 591.1566 0.0000 147.8350After that try 13 linear regression to reach better results
lm2_1=lm(net~trnd+w_day+mon+hour+prevday+prevweek,
tmp2)
summary(lm2_1)##
## Call:
## lm(formula = net ~ trnd + w_day + mon + hour + prevday + prevweek,
## data = tmp2)
##
## Residuals:
## Min 1Q Median 3Q Max
## -5876.5 -521.5 -10.9 500.2 6631.5
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) -7.709e+01 3.656e+01 -2.108 0.035014 *
## trnd 1.308e-02 7.617e-04 17.173 < 2e-16 ***
## w_dayCum 7.933e+01 2.050e+01 3.870 0.000109 ***
## w_dayÇar 1.143e+02 2.051e+01 5.575 2.50e-08 ***
## w_dayPaz -1.510e+02 2.056e+01 -7.346 2.10e-13 ***
## w_dayPer 1.679e+02 2.052e+01 8.182 2.93e-16 ***
## w_dayPts 2.393e+02 2.062e+01 11.606 < 2e-16 ***
## w_daySal 1.678e+02 2.051e+01 8.180 2.95e-16 ***
## monAra -1.854e+02 2.705e+01 -6.856 7.26e-12 ***
## monEki -1.600e+02 2.702e+01 -5.922 3.21e-09 ***
## monEyl -2.305e+02 2.724e+01 -8.460 < 2e-16 ***
## monHaz -8.766e+01 2.722e+01 -3.220 0.001283 **
## monKas -1.667e+02 2.727e+01 -6.113 9.90e-10 ***
## monMar -1.850e+02 2.713e+01 -6.817 9.51e-12 ***
## monMay -1.908e+02 2.717e+01 -7.021 2.25e-12 ***
## monNis -2.542e+02 2.743e+01 -9.269 < 2e-16 ***
## monOca -1.894e+02 2.583e+01 -7.330 2.36e-13 ***
## monŞub -1.405e+02 2.779e+01 -5.057 4.29e-07 ***
## monTem 7.102e+00 2.684e+01 0.265 0.791309
## hour1 7.730e+01 3.794e+01 2.038 0.041600 *
## hour2 4.389e+01 3.793e+01 1.157 0.247137
## hour3 1.599e+01 3.792e+01 0.422 0.673182
## hour4 -1.380e+01 3.792e+01 -0.364 0.715899
## hour5 -2.599e+01 3.792e+01 -0.685 0.493056
## hour6 -7.628e+01 3.794e+01 -2.011 0.044361 *
## hour7 -6.048e+01 3.793e+01 -1.594 0.110863
## hour8 2.604e+01 3.792e+01 0.687 0.492355
## hour9 1.562e+02 3.799e+01 4.112 3.94e-05 ***
## hour10 1.464e+02 3.798e+01 3.854 0.000116 ***
## hour11 2.139e+02 3.805e+01 5.621 1.91e-08 ***
## hour12 1.143e+02 3.796e+01 3.012 0.002598 **
## hour13 7.913e+01 3.794e+01 2.086 0.037009 *
## hour14 2.123e+02 3.805e+01 5.580 2.43e-08 ***
## hour15 2.076e+02 3.804e+01 5.456 4.92e-08 ***
## hour16 1.754e+02 3.801e+01 4.615 3.95e-06 ***
## hour17 2.205e+02 3.806e+01 5.793 7.01e-09 ***
## hour18 2.122e+02 3.805e+01 5.577 2.47e-08 ***
## hour19 1.985e+02 3.803e+01 5.220 1.80e-07 ***
## hour20 1.930e+02 3.803e+01 5.076 3.88e-07 ***
## hour21 2.063e+02 3.804e+01 5.423 5.90e-08 ***
## hour22 1.490e+02 3.798e+01 3.923 8.76e-05 ***
## hour23 1.108e+02 3.796e+01 2.919 0.003514 **
## prevday 4.179e-01 5.542e-03 75.405 < 2e-16 ***
## prevweek 7.769e-02 5.545e-03 14.011 < 2e-16 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 896.2 on 26764 degrees of freedom
## Multiple R-squared: 0.2737, Adjusted R-squared: 0.2726
## F-statistic: 234.6 on 43 and 26764 DF, p-value: < 2.2e-16tmp2[,predicted:=predict(lm2_1,tmp2)]
#head(tmp)
ggplot(tmp2[Date=='2019-05-08'] ,aes(x=datetime)) +
geom_line(aes(y=actual,color='real')) +
geom_line(aes(y=predicted,color='predicted'))
tmp2[,residual:=actual-predicted]
fit_res_tree=rpart(residual~w_day+mon+hour+prevday+prevweek+
loc1_DSWRF_surface+loc1_RH_2.m.above.ground+loc1_TCDC_low.cloud.layer+loc1_TMP_2.m.above.ground+loc1_wdir_10m+loc1_ws_10m+
loc2_DSWRF_surface+loc2_RH_2.m.above.ground+loc2_TCDC_low.cloud.layer+loc2_TMP_2.m.above.ground+loc2_wdir_10m+loc2_ws_10m+
loc3_DSWRF_surface+loc3_RH_2.m.above.ground+loc3_TCDC_low.cloud.layer+loc3_TMP_2.m.above.ground+loc3_wdir_10m+loc3_ws_10m+
loc4_DSWRF_surface+loc4_RH_2.m.above.ground+loc4_TCDC_low.cloud.layer+loc4_TMP_2.m.above.ground+loc4_wdir_10m+loc4_ws_10m+
loc5_DSWRF_surface+loc5_RH_2.m.above.ground+loc5_TCDC_low.cloud.layer+loc5_TMP_2.m.above.ground+loc5_wdir_10m+loc5_ws_10m+
loc6_DSWRF_surface+loc6_RH_2.m.above.ground+loc6_TCDC_low.cloud.layer+loc6_TMP_2.m.above.ground+loc6_wdir_10m+loc6_ws_10m+
loc7_DSWRF_surface+loc7_RH_2.m.above.ground+loc7_TCDC_low.cloud.layer+loc7_TMP_2.m.above.ground+loc7_wdir_10m+loc7_ws_10m,tmp2,
control=rpart.control(cp=0,maxdepth=3))
fancyRpartPlot(fit_res_tree)
tmp2[,loc7tmp2ag:=as.numeric(loc7_TMP_2.m.above.ground>=30)]
tmp2[,pv:=as.numeric(prevweek>=2070)]
tmp2[,loc7tmp2ag2:=as.numeric(loc7_TMP_2.m.above.ground>=31)]
lm2_2=lm(net~trnd+w_day+mon+hour+prevday+prevweek+loc7tmp2ag:pv:loc7tmp2ag2,
tmp2)
summary(lm2_2)##
## Call:
## lm(formula = net ~ trnd + w_day + mon + hour + prevday + prevweek +
## loc7tmp2ag:pv:loc7tmp2ag2, data = tmp2)
##
## Residuals:
## Min 1Q Median 3Q Max
## -5545.6 -516.0 -7.5 499.2 5789.6
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) -1.136e+02 3.596e+01 -3.159 0.001583 **
## trnd 1.324e-02 7.488e-04 17.679 < 2e-16 ***
## w_dayCum 7.455e+01 2.015e+01 3.699 0.000217 ***
## w_dayÇar 1.103e+02 2.016e+01 5.470 4.55e-08 ***
## w_dayPaz -1.566e+02 2.021e+01 -7.746 9.82e-15 ***
## w_dayPer 1.563e+02 2.017e+01 7.746 9.80e-15 ***
## w_dayPts 2.225e+02 2.028e+01 10.975 < 2e-16 ***
## w_daySal 1.651e+02 2.016e+01 8.186 2.82e-16 ***
## monAra -1.331e+02 2.664e+01 -4.994 5.94e-07 ***
## monEki -1.102e+02 2.661e+01 -4.140 3.48e-05 ***
## monEyl -1.827e+02 2.682e+01 -6.813 9.75e-12 ***
## monHaz -3.834e+01 2.681e+01 -1.430 0.152707
## monKas -1.165e+02 2.686e+01 -4.339 1.43e-05 ***
## monMar -1.394e+02 2.671e+01 -5.217 1.83e-07 ***
## monMay -1.485e+02 2.674e+01 -5.553 2.84e-08 ***
## monNis -2.128e+02 2.700e+01 -7.884 3.29e-15 ***
## monOca -1.443e+02 2.544e+01 -5.673 1.42e-08 ***
## monŞub -9.414e+01 2.736e+01 -3.441 0.000580 ***
## monTem 6.196e+01 2.644e+01 2.343 0.019137 *
## hour1 8.299e+01 3.729e+01 2.225 0.026077 *
## hour2 4.712e+01 3.728e+01 1.264 0.206257
## hour3 1.719e+01 3.728e+01 0.461 0.644731
## hour4 -1.480e+01 3.728e+01 -0.397 0.691351
## hour5 -2.790e+01 3.728e+01 -0.748 0.454281
## hour6 -8.191e+01 3.729e+01 -2.196 0.028067 *
## hour7 -6.490e+01 3.729e+01 -1.741 0.081767 .
## hour8 2.794e+01 3.728e+01 0.750 0.453555
## hour9 1.677e+02 3.735e+01 4.490 7.16e-06 ***
## hour10 1.572e+02 3.734e+01 4.209 2.57e-05 ***
## hour11 2.209e+02 3.741e+01 5.906 3.55e-09 ***
## hour12 1.140e+02 3.731e+01 3.055 0.002250 **
## hour13 7.622e+01 3.729e+01 2.044 0.040989 *
## hour14 2.105e+02 3.740e+01 5.627 1.85e-08 ***
## hour15 1.966e+02 3.740e+01 5.258 1.47e-07 ***
## hour16 1.665e+02 3.737e+01 4.457 8.36e-06 ***
## hour17 2.193e+02 3.742e+01 5.861 4.66e-09 ***
## hour18 2.191e+02 3.740e+01 5.858 4.75e-09 ***
## hour19 2.088e+02 3.739e+01 5.584 2.37e-08 ***
## hour20 1.985e+02 3.738e+01 5.310 1.10e-07 ***
## hour21 2.128e+02 3.740e+01 5.689 1.29e-08 ***
## hour22 1.599e+02 3.734e+01 4.283 1.85e-05 ***
## hour23 1.189e+02 3.731e+01 3.187 0.001441 **
## prevday 3.916e-01 5.515e-03 71.004 < 2e-16 ***
## prevweek 6.677e-02 5.463e-03 12.222 < 2e-16 ***
## loc7tmp2ag:pv:loc7tmp2ag2 4.872e+03 1.594e+02 30.561 < 2e-16 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 880.9 on 26763 degrees of freedom
## Multiple R-squared: 0.2982, Adjusted R-squared: 0.2971
## F-statistic: 258.5 on 44 and 26763 DF, p-value: < 2.2e-16tmp2[,predicted:=predict(lm2_2,tmp2)]
#head(tmp)
ggplot(tmp2[Date=='2019-05-08'] ,aes(x=datetime)) +
geom_line(aes(y=actual,color='real')) +
geom_line(aes(y=predicted,color='predicted'))
tmp2[,residual:=actual-predicted]
fit_res_tree=rpart(residual~w_day+mon+hour+prevday+prevweek+
loc1_DSWRF_surface+loc1_RH_2.m.above.ground+loc1_TCDC_low.cloud.layer+loc1_TMP_2.m.above.ground+loc1_wdir_10m+loc1_ws_10m+
loc2_DSWRF_surface+loc2_RH_2.m.above.ground+loc2_TCDC_low.cloud.layer+loc2_TMP_2.m.above.ground+loc2_wdir_10m+loc2_ws_10m+
loc3_DSWRF_surface+loc3_RH_2.m.above.ground+loc3_TCDC_low.cloud.layer+loc3_TMP_2.m.above.ground+loc3_wdir_10m+loc3_ws_10m+
loc4_DSWRF_surface+loc4_RH_2.m.above.ground+loc4_TCDC_low.cloud.layer+loc4_TMP_2.m.above.ground+loc4_wdir_10m+loc4_ws_10m+
loc5_DSWRF_surface+loc5_RH_2.m.above.ground+loc5_TCDC_low.cloud.layer+loc5_TMP_2.m.above.ground+loc5_wdir_10m+loc5_ws_10m+
loc6_DSWRF_surface+loc6_RH_2.m.above.ground+loc6_TCDC_low.cloud.layer+loc6_TMP_2.m.above.ground+loc6_wdir_10m+loc6_ws_10m+
loc7_DSWRF_surface+loc7_RH_2.m.above.ground+loc7_TCDC_low.cloud.layer+loc7_wdir_10m+loc7_ws_10m,tmp2,
control=rpart.control(cp=0,maxdepth=3))
fancyRpartPlot(fit_res_tree)
tmp2[,loc5tmp2ag:=as.numeric(loc5_TMP_2.m.above.ground>=29)]
tmp2[,pd:=as.numeric(prevday>=2329)]
tmp2[,workday:=as.numeric(!w_day %in% c('Cmt','Paz'))]
lm2_3=lm(net~trnd+w_day+mon+hour+prevday+prevweek+loc7tmp2ag:pv:loc7tmp2ag2+loc5tmp2ag:pd:workday,
tmp2)
summary(lm2_3)##
## Call:
## lm(formula = net ~ trnd + w_day + mon + hour + prevday + prevweek +
## loc7tmp2ag:pv:loc7tmp2ag2 + loc5tmp2ag:pd:workday, data = tmp2)
##
## Residuals:
## Min 1Q Median 3Q Max
## -5168.1 -512.6 -6.8 497.3 5811.9
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) -1.367e+02 3.561e+01 -3.838 0.000124 ***
## trnd 1.325e-02 7.411e-04 17.885 < 2e-16 ***
## w_dayCum 6.758e+01 1.995e+01 3.388 0.000705 ***
## w_dayÇar 1.093e+02 1.995e+01 5.480 4.30e-08 ***
## w_dayPaz -1.485e+02 2.001e+01 -7.421 1.20e-13 ***
## w_dayPer 1.542e+02 1.996e+01 7.724 1.16e-14 ***
## w_dayPts 2.239e+02 2.007e+01 11.160 < 2e-16 ***
## w_daySal 1.685e+02 1.996e+01 8.442 < 2e-16 ***
## monAra -9.618e+01 2.641e+01 -3.641 0.000272 ***
## monEki -7.438e+01 2.638e+01 -2.819 0.004814 **
## monEyl -1.691e+02 2.655e+01 -6.369 1.93e-10 ***
## monHaz -1.134e+01 2.656e+01 -0.427 0.669421
## monKas -8.077e+01 2.662e+01 -3.034 0.002417 **
## monMar -1.087e+02 2.647e+01 -4.105 4.06e-05 ***
## monMay -1.200e+02 2.650e+01 -4.529 5.96e-06 ***
## monNis -1.864e+02 2.674e+01 -6.972 3.20e-12 ***
## monOca -1.136e+02 2.521e+01 -4.506 6.62e-06 ***
## monŞub -6.207e+01 2.711e+01 -2.290 0.022055 *
## monTem 6.308e+01 2.617e+01 2.410 0.015956 *
## hour1 8.814e+01 3.691e+01 2.388 0.016953 *
## hour2 5.005e+01 3.690e+01 1.356 0.174978
## hour3 1.826e+01 3.689e+01 0.495 0.620646
## hour4 -1.571e+01 3.689e+01 -0.426 0.670167
## hour5 -2.962e+01 3.689e+01 -0.803 0.422043
## hour6 -8.700e+01 3.691e+01 -2.357 0.018429 *
## hour7 -6.892e+01 3.690e+01 -1.868 0.061823 .
## hour8 2.968e+01 3.689e+01 0.804 0.421157
## hour9 1.781e+02 3.697e+01 4.818 1.46e-06 ***
## hour10 1.654e+02 3.696e+01 4.474 7.70e-06 ***
## hour11 2.208e+02 3.702e+01 5.965 2.47e-09 ***
## hour12 1.057e+02 3.693e+01 2.863 0.004203 **
## hour13 5.769e+01 3.692e+01 1.563 0.118124
## hour14 1.865e+02 3.703e+01 5.036 4.78e-07 ***
## hour15 1.738e+02 3.703e+01 4.694 2.69e-06 ***
## hour16 1.440e+02 3.699e+01 3.892 9.96e-05 ***
## hour17 1.943e+02 3.705e+01 5.244 1.58e-07 ***
## hour18 2.142e+02 3.702e+01 5.786 7.30e-09 ***
## hour19 2.070e+02 3.701e+01 5.593 2.25e-08 ***
## hour20 2.129e+02 3.700e+01 5.753 8.88e-09 ***
## hour21 2.280e+02 3.702e+01 6.159 7.44e-10 ***
## hour22 1.699e+02 3.696e+01 4.596 4.32e-06 ***
## hour23 1.263e+02 3.693e+01 3.420 0.000627 ***
## prevday 3.562e-01 5.660e-03 62.929 < 2e-16 ***
## prevweek 6.855e-02 5.407e-03 12.677 < 2e-16 ***
## loc7tmp2ag:pv:loc7tmp2ag2 4.055e+03 1.615e+02 25.109 < 2e-16 ***
## loc5tmp2ag:pd:workday 1.764e+03 7.454e+01 23.666 < 2e-16 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 871.9 on 26762 degrees of freedom
## Multiple R-squared: 0.3126, Adjusted R-squared: 0.3115
## F-statistic: 270.5 on 45 and 26762 DF, p-value: < 2.2e-16tmp2[,predicted:=predict(lm2_3,tmp2)]
#head(tmp)
ggplot(tmp2[Date=='2019-05-08'] ,aes(x=datetime)) +
geom_line(aes(y=actual,color='real')) +
geom_line(aes(y=predicted,color='predicted'))
tmp2[,residual:=actual-predicted]
fit_res_tree=rpart(residual~mon+hour+
loc1_DSWRF_surface+loc1_RH_2.m.above.ground+loc1_TCDC_low.cloud.layer+loc1_TMP_2.m.above.ground+loc1_wdir_10m+loc1_ws_10m+
loc2_DSWRF_surface+loc2_RH_2.m.above.ground+loc2_TCDC_low.cloud.layer+loc2_TMP_2.m.above.ground+loc2_wdir_10m+loc2_ws_10m+
loc3_DSWRF_surface+loc3_RH_2.m.above.ground+loc3_TCDC_low.cloud.layer+loc3_TMP_2.m.above.ground+loc3_wdir_10m+loc3_ws_10m+
loc4_DSWRF_surface+loc4_RH_2.m.above.ground+loc4_TCDC_low.cloud.layer+loc4_TMP_2.m.above.ground+loc4_wdir_10m+loc4_ws_10m+
loc5_DSWRF_surface+loc5_RH_2.m.above.ground+loc5_TCDC_low.cloud.layer+loc5_wdir_10m+loc5_ws_10m+
loc6_DSWRF_surface+loc6_RH_2.m.above.ground+loc6_TCDC_low.cloud.layer+loc6_TMP_2.m.above.ground+loc6_wdir_10m+loc6_ws_10m+
loc7_DSWRF_surface+loc7_RH_2.m.above.ground+loc7_TCDC_low.cloud.layer+loc7_wdir_10m+loc7_ws_10m,tmp2,
control=rpart.control(cp=0,maxdepth=3))
fancyRpartPlot(fit_res_tree)
tmp2[,loc4tmp2ag:=as.numeric(loc4_TMP_2.m.above.ground<8.6)]
tmp2[,whour:=as.numeric(as.numeric(hour)<9 | as.numeric(hour)>18)]
tmp2[,loc3wdir:=as.numeric(loc3_wdir_10m<336)]
lm2_4=lm(net~trnd+w_day+mon+hour+prevday+prevweek+loc7tmp2ag:pv:loc7tmp2ag2+loc5tmp2ag:pd:workday+loc4tmp2ag:whour:loc3wdir,
tmp2)
summary(lm2_4)##
## Call:
## lm(formula = net ~ trnd + w_day + mon + hour + prevday + prevweek +
## loc7tmp2ag:pv:loc7tmp2ag2 + loc5tmp2ag:pd:workday + loc4tmp2ag:whour:loc3wdir,
## data = tmp2)
##
## Residuals:
## Min 1Q Median 3Q Max
## -5167.8 -508.3 -8.1 496.9 5795.6
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) -1.459e+02 3.559e+01 -4.099 4.16e-05 ***
## trnd 1.318e-02 7.404e-04 17.805 < 2e-16 ***
## w_dayCum 6.535e+01 1.993e+01 3.279 0.001043 **
## w_dayÇar 1.099e+02 1.993e+01 5.514 3.54e-08 ***
## w_dayPaz -1.487e+02 1.999e+01 -7.438 1.05e-13 ***
## w_dayPer 1.529e+02 1.995e+01 7.667 1.83e-14 ***
## w_dayPts 2.244e+02 2.005e+01 11.195 < 2e-16 ***
## w_daySal 1.711e+02 1.994e+01 8.581 < 2e-16 ***
## monAra -1.263e+02 2.671e+01 -4.727 2.29e-06 ***
## monEki -7.533e+01 2.635e+01 -2.859 0.004259 **
## monEyl -1.702e+02 2.653e+01 -6.415 1.43e-10 ***
## monHaz -1.262e+01 2.653e+01 -0.476 0.634319
## monKas -9.099e+01 2.663e+01 -3.416 0.000635 ***
## monMar -1.408e+02 2.681e+01 -5.253 1.51e-07 ***
## monMay -1.217e+02 2.647e+01 -4.596 4.32e-06 ***
## monNis -2.009e+02 2.679e+01 -7.499 6.65e-14 ***
## monOca -1.625e+02 2.606e+01 -6.235 4.60e-10 ***
## monŞub -1.059e+02 2.774e+01 -3.816 0.000136 ***
## monTem 6.290e+01 2.615e+01 2.406 0.016148 *
## hour1 8.743e+01 3.688e+01 2.371 0.017755 *
## hour2 4.877e+01 3.686e+01 1.323 0.185807
## hour3 1.483e+01 3.686e+01 0.402 0.687405
## hour4 -1.996e+01 3.686e+01 -0.541 0.588249
## hour5 -3.595e+01 3.687e+01 -0.975 0.329594
## hour6 -9.451e+01 3.689e+01 -2.562 0.010416 *
## hour7 -7.770e+01 3.689e+01 -2.106 0.035180 *
## hour8 5.609e+01 3.704e+01 1.514 0.129920
## hour9 2.051e+02 3.712e+01 5.526 3.30e-08 ***
## hour10 1.923e+02 3.711e+01 5.184 2.19e-07 ***
## hour11 2.482e+02 3.718e+01 6.677 2.49e-11 ***
## hour12 1.326e+02 3.708e+01 3.577 0.000349 ***
## hour13 8.445e+01 3.706e+01 2.278 0.022706 *
## hour14 2.140e+02 3.719e+01 5.753 8.84e-09 ***
## hour15 2.012e+02 3.718e+01 5.412 6.29e-08 ***
## hour16 1.712e+02 3.715e+01 4.610 4.05e-06 ***
## hour17 2.218e+02 3.720e+01 5.961 2.53e-09 ***
## hour18 2.312e+02 3.706e+01 6.238 4.49e-10 ***
## hour19 2.189e+02 3.701e+01 5.915 3.36e-09 ***
## hour20 2.212e+02 3.698e+01 5.980 2.25e-09 ***
## hour21 2.345e+02 3.699e+01 6.338 2.36e-10 ***
## hour22 1.743e+02 3.693e+01 4.721 2.36e-06 ***
## hour23 1.280e+02 3.689e+01 3.470 0.000521 ***
## prevday 3.548e-01 5.658e-03 62.712 < 2e-16 ***
## prevweek 6.753e-02 5.404e-03 12.498 < 2e-16 ***
## loc7tmp2ag:pv:loc7tmp2ag2 4.058e+03 1.613e+02 25.150 < 2e-16 ***
## loc5tmp2ag:pd:workday 1.757e+03 7.447e+01 23.594 < 2e-16 ***
## loc4tmp2ag:whour:loc3wdir 1.506e+02 2.065e+01 7.292 3.13e-13 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 871 on 26761 degrees of freedom
## Multiple R-squared: 0.314, Adjusted R-squared: 0.3128
## F-statistic: 266.3 on 46 and 26761 DF, p-value: < 2.2e-16tmp2[,predicted:=predict(lm2_4,tmp2)]
#head(tmp)
ggplot(tmp2[Date=='2019-05-08'] ,aes(x=datetime)) +
geom_line(aes(y=actual,color='real')) +
geom_line(aes(y=predicted,color='predicted'))
tmp2[,residual:=actual-predicted]
fit_res_tree=rpart(residual~mon+
loc1_DSWRF_surface+loc1_RH_2.m.above.ground+loc1_TCDC_low.cloud.layer+loc1_TMP_2.m.above.ground+loc1_wdir_10m+loc1_ws_10m+
loc2_DSWRF_surface+loc2_RH_2.m.above.ground+loc2_TCDC_low.cloud.layer+loc2_TMP_2.m.above.ground+loc2_wdir_10m+loc2_ws_10m+
loc3_DSWRF_surface+loc3_RH_2.m.above.ground+loc3_TCDC_low.cloud.layer+loc3_TMP_2.m.above.ground+loc3_ws_10m+
loc4_DSWRF_surface+loc4_RH_2.m.above.ground+loc4_TCDC_low.cloud.layer+loc4_wdir_10m+loc4_ws_10m+
loc5_DSWRF_surface+loc5_RH_2.m.above.ground+loc5_TCDC_low.cloud.layer+loc5_wdir_10m+loc5_ws_10m+
loc6_DSWRF_surface+loc6_RH_2.m.above.ground+loc6_TCDC_low.cloud.layer+loc6_TMP_2.m.above.ground+loc6_wdir_10m+loc6_ws_10m+
loc7_DSWRF_surface+loc7_RH_2.m.above.ground+loc7_TCDC_low.cloud.layer+loc7_wdir_10m+loc7_ws_10m,tmp2,
control=rpart.control(cp=0,maxdepth=3))
fancyRpartPlot(fit_res_tree)
tmp2[,loc6tmp2ag:=as.numeric(loc6_TMP_2.m.above.ground>=26)]
tmp2[,loc7ws:=as.numeric(loc7_ws_10m<3)]
tmp2[,loc1wdir:=as.numeric(loc1_wdir_10m>=172)]
lm2_5=lm(net~trnd+w_day+mon+hour+prevday+prevweek+loc7tmp2ag:pv:loc7tmp2ag2+loc5tmp2ag:pd:workday+loc4tmp2ag:whour:loc3wdir+loc6tmp2ag:loc7ws:loc1wdir,
tmp2)
summary(lm2_5)##
## Call:
## lm(formula = net ~ trnd + w_day + mon + hour + prevday + prevweek +
## loc7tmp2ag:pv:loc7tmp2ag2 + loc5tmp2ag:pd:workday + loc4tmp2ag:whour:loc3wdir +
## loc6tmp2ag:loc7ws:loc1wdir, data = tmp2)
##
## Residuals:
## Min 1Q Median 3Q Max
## -5121.1 -509.0 -7.6 497.5 5052.3
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) -1.542e+02 3.545e+01 -4.350 1.37e-05 ***
## trnd 1.295e-02 7.375e-04 17.553 < 2e-16 ***
## w_dayCum 6.639e+01 1.985e+01 3.345 0.000825 ***
## w_dayÇar 1.115e+02 1.985e+01 5.615 1.98e-08 ***
## w_dayPaz -1.499e+02 1.991e+01 -7.532 5.17e-14 ***
## w_dayPer 1.538e+02 1.986e+01 7.744 9.99e-15 ***
## w_dayPts 2.194e+02 1.997e+01 10.985 < 2e-16 ***
## w_daySal 1.679e+02 1.986e+01 8.455 < 2e-16 ***
## monAra -1.079e+02 2.663e+01 -4.051 5.12e-05 ***
## monEki -6.422e+01 2.626e+01 -2.446 0.014459 *
## monEyl -1.528e+02 2.644e+01 -5.779 7.60e-09 ***
## monHaz 1.097e+00 2.644e+01 0.041 0.966911
## monKas -7.347e+01 2.655e+01 -2.767 0.005661 **
## monMar -1.246e+02 2.672e+01 -4.662 3.14e-06 ***
## monMay -1.096e+02 2.637e+01 -4.156 3.25e-05 ***
## monNis -1.855e+02 2.670e+01 -6.948 3.78e-12 ***
## monOca -1.455e+02 2.598e+01 -5.601 2.15e-08 ***
## monŞub -8.946e+01 2.765e+01 -3.236 0.001215 **
## monTem 6.803e+01 2.604e+01 2.612 0.009002 **
## hour1 8.800e+01 3.672e+01 2.396 0.016572 *
## hour2 4.913e+01 3.671e+01 1.338 0.180857
## hour3 1.505e+01 3.671e+01 0.410 0.681823
## hour4 -1.993e+01 3.671e+01 -0.543 0.587265
## hour5 -3.594e+01 3.672e+01 -0.979 0.327656
## hour6 -9.483e+01 3.674e+01 -2.581 0.009850 **
## hour7 -7.787e+01 3.674e+01 -2.120 0.034045 *
## hour8 5.550e+01 3.688e+01 1.505 0.132447
## hour9 2.029e+02 3.696e+01 5.489 4.09e-08 ***
## hour10 1.812e+02 3.696e+01 4.903 9.49e-07 ***
## hour11 2.394e+02 3.703e+01 6.465 1.03e-10 ***
## hour12 1.143e+02 3.695e+01 3.094 0.001980 **
## hour13 6.783e+01 3.693e+01 1.837 0.066272 .
## hour14 2.028e+02 3.704e+01 5.475 4.42e-08 ***
## hour15 1.971e+02 3.703e+01 5.322 1.03e-07 ***
## hour16 1.676e+02 3.699e+01 4.529 5.94e-06 ***
## hour17 2.198e+02 3.705e+01 5.932 3.04e-09 ***
## hour18 2.320e+02 3.691e+01 6.287 3.29e-10 ***
## hour19 2.203e+02 3.685e+01 5.979 2.28e-09 ***
## hour20 2.230e+02 3.683e+01 6.054 1.43e-09 ***
## hour21 2.364e+02 3.684e+01 6.417 1.41e-10 ***
## hour22 1.740e+02 3.678e+01 4.731 2.24e-06 ***
## hour23 1.288e+02 3.674e+01 3.505 0.000458 ***
## prevday 3.520e-01 5.638e-03 62.441 < 2e-16 ***
## prevweek 6.678e-02 5.382e-03 12.409 < 2e-16 ***
## loc7tmp2ag:pv:loc7tmp2ag2 3.692e+03 1.626e+02 22.712 < 2e-16 ***
## loc5tmp2ag:pd:workday 1.609e+03 7.483e+01 21.506 < 2e-16 ***
## loc4tmp2ag:whour:loc3wdir 1.461e+02 2.057e+01 7.105 1.23e-12 ***
## loc6tmp2ag:loc7ws:loc1wdir 1.428e+03 9.591e+01 14.889 < 2e-16 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 867.5 on 26760 degrees of freedom
## Multiple R-squared: 0.3196, Adjusted R-squared: 0.3184
## F-statistic: 267.5 on 47 and 26760 DF, p-value: < 2.2e-16tmp2[,predicted:=predict(lm2_5,tmp2)]
#head(tmp)
ggplot(tmp2[Date=='2019-05-08'] ,aes(x=datetime)) +
geom_line(aes(y=actual,color='real')) +
geom_line(aes(y=predicted,color='predicted'))
tmp2[,residual:=actual-predicted]
fit_res_tree=rpart(residual~mon+
loc1_DSWRF_surface+loc1_RH_2.m.above.ground+loc1_TCDC_low.cloud.layer+loc1_TMP_2.m.above.ground+loc1_ws_10m+
loc2_DSWRF_surface+loc2_RH_2.m.above.ground+loc2_TCDC_low.cloud.layer+loc2_TMP_2.m.above.ground+loc2_wdir_10m+loc2_ws_10m+
loc3_DSWRF_surface+loc3_RH_2.m.above.ground+loc3_TCDC_low.cloud.layer+loc3_TMP_2.m.above.ground+loc3_ws_10m+
loc4_DSWRF_surface+loc4_RH_2.m.above.ground+loc4_TCDC_low.cloud.layer+loc4_wdir_10m+loc4_ws_10m+
loc5_DSWRF_surface+loc5_RH_2.m.above.ground+loc5_TCDC_low.cloud.layer+loc5_wdir_10m+loc5_ws_10m+
loc6_DSWRF_surface+loc6_RH_2.m.above.ground+loc6_TCDC_low.cloud.layer+loc6_wdir_10m+loc6_ws_10m+
loc7_DSWRF_surface+loc7_RH_2.m.above.ground+loc7_TCDC_low.cloud.layer+loc7_wdir_10m,tmp2,
control=rpart.control(cp=0,maxdepth=3))
fancyRpartPlot(fit_res_tree)
tmp2[,loc4wdir:=as.numeric(loc4_wdir_10m<170)]
tmp2[,loc7wdir:=as.numeric(loc7_wdir_10m>=214)]
tmp2[,loc3tmp2ag:=as.numeric(loc3_TMP_2.m.above.ground>=20)]
lm2_6=lm(net~trnd+w_day+mon+hour+prevday+prevweek+loc7tmp2ag:pv:loc7tmp2ag2+loc5tmp2ag:pd:workday+loc4tmp2ag:whour:loc3wdir+loc6tmp2ag:loc7ws:loc1wdir
+loc4wdir:loc7wdir+loc3tmp2ag,
tmp2)
summary(lm2_6)##
## Call:
## lm(formula = net ~ trnd + w_day + mon + hour + prevday + prevweek +
## loc7tmp2ag:pv:loc7tmp2ag2 + loc5tmp2ag:pd:workday + loc4tmp2ag:whour:loc3wdir +
## loc6tmp2ag:loc7ws:loc1wdir + loc4wdir:loc7wdir + loc3tmp2ag,
## data = tmp2)
##
## Residuals:
## Min 1Q Median 3Q Max
## -4970.8 -512.5 -4.6 497.1 5010.2
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) -3.212e+02 3.670e+01 -8.751 < 2e-16 ***
## trnd 1.380e-02 7.335e-04 18.808 < 2e-16 ***
## w_dayCum 6.751e+01 1.970e+01 3.427 0.000611 ***
## w_dayÇar 1.033e+02 1.970e+01 5.245 1.58e-07 ***
## w_dayPaz -1.541e+02 1.976e+01 -7.799 6.48e-15 ***
## w_dayPer 1.520e+02 1.971e+01 7.709 1.31e-14 ***
## w_dayPts 2.138e+02 1.982e+01 10.784 < 2e-16 ***
## w_daySal 1.556e+02 1.972e+01 7.889 3.16e-15 ***
## monAra 7.151e+01 2.923e+01 2.446 0.014448 *
## monEki 4.220e+01 2.758e+01 1.530 0.125971
## monEyl -1.025e+02 2.665e+01 -3.848 0.000119 ***
## monHaz 6.364e+01 2.652e+01 2.399 0.016428 *
## monKas 9.300e+01 2.897e+01 3.210 0.001330 **
## monMar 5.181e+01 2.936e+01 1.765 0.077643 .
## monMay -3.200e+01 2.695e+01 -1.188 0.234986
## monNis -3.963e+01 2.888e+01 -1.372 0.170055
## monOca 4.075e+01 2.882e+01 1.414 0.157310
## monŞub 9.726e+01 3.032e+01 3.207 0.001342 **
## monTem 5.717e+01 2.585e+01 2.211 0.027026 *
## hour1 1.010e+02 3.645e+01 2.770 0.005608 **
## hour2 6.896e+01 3.645e+01 1.892 0.058505 .
## hour3 4.178e+01 3.646e+01 1.146 0.251839
## hour4 1.075e+01 3.647e+01 0.295 0.768287
## hour5 5.069e+00 3.650e+01 0.139 0.889560
## hour6 -5.554e+01 3.652e+01 -1.521 0.128338
## hour7 -4.368e+01 3.649e+01 -1.197 0.231396
## hour8 6.874e+01 3.663e+01 1.877 0.060570 .
## hour9 2.000e+02 3.674e+01 5.443 5.30e-08 ***
## hour10 1.556e+02 3.681e+01 4.228 2.37e-05 ***
## hour11 1.999e+02 3.695e+01 5.410 6.35e-08 ***
## hour12 5.801e+01 3.697e+01 1.569 0.116644
## hour13 4.936e-01 3.703e+01 0.013 0.989366
## hour14 1.370e+02 3.715e+01 3.687 0.000227 ***
## hour15 1.278e+02 3.715e+01 3.441 0.000581 ***
## hour16 9.723e+01 3.711e+01 2.620 0.008802 **
## hour17 1.458e+02 3.715e+01 3.924 8.74e-05 ***
## hour18 1.592e+02 3.691e+01 4.313 1.61e-05 ***
## hour19 1.490e+02 3.679e+01 4.050 5.13e-05 ***
## hour20 1.648e+02 3.667e+01 4.495 6.99e-06 ***
## hour21 1.962e+02 3.661e+01 5.358 8.46e-08 ***
## hour22 1.465e+02 3.652e+01 4.011 6.06e-05 ***
## hour23 1.204e+02 3.646e+01 3.303 0.000958 ***
## prevday 3.400e-01 5.641e-03 60.270 < 2e-16 ***
## prevweek 6.358e-02 5.345e-03 11.896 < 2e-16 ***
## loc3tmp2ag 2.022e+02 1.846e+01 10.959 < 2e-16 ***
## loc4wdir:loc7wdir 2.809e+02 1.714e+01 16.386 < 2e-16 ***
## loc7tmp2ag:pv:loc7tmp2ag2 3.702e+03 1.613e+02 22.945 < 2e-16 ***
## loc5tmp2ag:pd:workday 1.570e+03 7.428e+01 21.133 < 2e-16 ***
## loc4tmp2ag:whour:loc3wdir 1.209e+02 2.061e+01 5.869 4.45e-09 ***
## loc6tmp2ag:loc7ws:loc1wdir 1.409e+03 9.525e+01 14.798 < 2e-16 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 860.8 on 26758 degrees of freedom
## Multiple R-squared: 0.33, Adjusted R-squared: 0.3288
## F-statistic: 269 on 49 and 26758 DF, p-value: < 2.2e-16tmp2[,predicted:=predict(lm2_6,tmp2)]
#head(tmp)
ggplot(tmp2[Date=='2019-05-08'] ,aes(x=datetime)) +
geom_line(aes(y=actual,color='real')) +
geom_line(aes(y=predicted,color='predicted'))
tmp2[,residual:=actual-predicted]
fit_res_tree=rpart(residual~mon+
loc1_DSWRF_surface+loc1_RH_2.m.above.ground+loc1_TCDC_low.cloud.layer+loc1_TMP_2.m.above.ground+loc1_ws_10m+
loc2_DSWRF_surface+loc2_RH_2.m.above.ground+loc2_TCDC_low.cloud.layer+loc2_TMP_2.m.above.ground+loc2_wdir_10m+loc2_ws_10m+
loc3_DSWRF_surface+loc3_RH_2.m.above.ground+loc3_TCDC_low.cloud.layer+loc3_ws_10m+
loc4_DSWRF_surface+loc4_RH_2.m.above.ground+loc4_TCDC_low.cloud.layer+loc4_ws_10m+
loc5_DSWRF_surface+loc5_RH_2.m.above.ground+loc5_TCDC_low.cloud.layer+loc5_wdir_10m+loc5_ws_10m+
loc6_DSWRF_surface+loc6_RH_2.m.above.ground+loc6_TCDC_low.cloud.layer+loc6_wdir_10m+loc6_ws_10m+
loc7_DSWRF_surface+loc7_RH_2.m.above.ground+loc7_TCDC_low.cloud.layer,tmp2,
control=rpart.control(cp=0,maxdepth=3))
fancyRpartPlot(fit_res_tree)
tmp2[,loc3tcdc:=as.numeric(loc3_TCDC_low.cloud.layer>=8.6)]
tmp2[,mon1:=as.numeric(!mon %in% c('Haz','Tem','Ağu','Eyl','Eki'))]
tmp2[,loc4rh:=as.numeric(loc4_RH_2.m.above.ground<69)]
lm2_7=lm(net~trnd+w_day+mon+hour+prevday+prevweek+loc7tmp2ag:pv:loc7tmp2ag2+loc5tmp2ag:pd:workday+loc4tmp2ag:whour:loc3wdir+loc6tmp2ag:loc7ws:loc1wdir
+loc4wdir:loc7wdir+loc3tmp2ag+loc3tcdc:mon1:loc4rh,
tmp2)
summary(lm2_7)##
## Call:
## lm(formula = net ~ trnd + w_day + mon + hour + prevday + prevweek +
## loc7tmp2ag:pv:loc7tmp2ag2 + loc5tmp2ag:pd:workday + loc4tmp2ag:whour:loc3wdir +
## loc6tmp2ag:loc7ws:loc1wdir + loc4wdir:loc7wdir + loc3tmp2ag +
## loc3tcdc:mon1:loc4rh, data = tmp2)
##
## Residuals:
## Min 1Q Median 3Q Max
## -4954.3 -507.6 -3.8 493.2 4819.8
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) -3.348e+02 3.635e+01 -9.209 < 2e-16 ***
## trnd 1.427e-02 7.267e-04 19.638 < 2e-16 ***
## w_dayCum 5.539e+01 1.951e+01 2.838 0.004536 **
## w_dayÇar 9.369e+01 1.952e+01 4.800 1.59e-06 ***
## w_dayPaz -1.543e+02 1.957e+01 -7.889 3.17e-15 ***
## w_dayPer 1.405e+02 1.953e+01 7.195 6.41e-13 ***
## w_dayPts 2.160e+02 1.963e+01 11.004 < 2e-16 ***
## w_daySal 1.555e+02 1.953e+01 7.964 1.73e-15 ***
## monAra 6.120e+01 2.895e+01 2.114 0.034541 *
## monEki 7.554e+01 2.735e+01 2.762 0.005746 **
## monEyl -8.632e+01 2.640e+01 -3.270 0.001077 **
## monHaz 7.978e+01 2.628e+01 3.036 0.002397 **
## monKas 8.858e+01 2.869e+01 3.087 0.002021 **
## monMar -1.217e+00 2.917e+01 -0.042 0.966721
## monMay -6.705e+01 2.673e+01 -2.509 0.012129 *
## monNis -1.154e+02 2.879e+01 -4.007 6.18e-05 ***
## monOca 3.712e+01 2.854e+01 1.301 0.193375
## monŞub 7.913e+01 3.004e+01 2.634 0.008442 **
## monTem 5.593e+01 2.560e+01 2.185 0.028920 *
## hour1 1.039e+02 3.610e+01 2.878 0.004006 **
## hour2 7.501e+01 3.610e+01 2.078 0.037714 *
## hour3 5.233e+01 3.611e+01 1.449 0.147333
## hour4 2.630e+01 3.613e+01 0.728 0.466547
## hour5 2.223e+01 3.616e+01 0.615 0.538612
## hour6 -3.988e+01 3.617e+01 -1.102 0.270266
## hour7 -3.044e+01 3.615e+01 -0.842 0.399718
## hour8 6.724e+01 3.628e+01 1.854 0.063806 .
## hour9 1.829e+02 3.640e+01 5.025 5.08e-07 ***
## hour10 1.235e+02 3.648e+01 3.385 0.000712 ***
## hour11 1.442e+02 3.667e+01 3.933 8.41e-05 ***
## hour12 -1.827e+01 3.676e+01 -0.497 0.619150
## hour13 -8.566e+01 3.686e+01 -2.324 0.020148 *
## hour14 4.608e+01 3.700e+01 1.245 0.212964
## hour15 2.940e+01 3.704e+01 0.794 0.427371
## hour16 4.964e+00 3.697e+01 0.134 0.893200
## hour17 5.354e+01 3.701e+01 1.447 0.147979
## hour18 8.280e+01 3.671e+01 2.256 0.024094 *
## hour19 8.640e+01 3.653e+01 2.365 0.018037 *
## hour20 1.185e+02 3.637e+01 3.259 0.001120 **
## hour21 1.625e+02 3.629e+01 4.479 7.53e-06 ***
## hour22 1.377e+02 3.617e+01 3.807 0.000141 ***
## hour23 1.175e+02 3.611e+01 3.255 0.001136 **
## prevday 3.316e-01 5.598e-03 59.232 < 2e-16 ***
## prevweek 6.486e-02 5.293e-03 12.253 < 2e-16 ***
## loc3tmp2ag 2.664e+02 1.849e+01 14.410 < 2e-16 ***
## loc4wdir:loc7wdir 2.978e+02 1.699e+01 17.524 < 2e-16 ***
## loc7tmp2ag:pv:loc7tmp2ag2 3.727e+03 1.598e+02 23.326 < 2e-16 ***
## loc5tmp2ag:pd:workday 1.626e+03 7.360e+01 22.094 < 2e-16 ***
## loc4tmp2ag:whour:loc3wdir 9.385e+01 2.044e+01 4.591 4.44e-06 ***
## loc6tmp2ag:loc7ws:loc1wdir 1.424e+03 9.433e+01 15.092 < 2e-16 ***
## loc3tcdc:mon1:loc4rh 4.396e+02 1.915e+01 22.952 < 2e-16 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 852.5 on 26757 degrees of freedom
## Multiple R-squared: 0.343, Adjusted R-squared: 0.3418
## F-statistic: 279.4 on 50 and 26757 DF, p-value: < 2.2e-16tmp2[,predicted:=predict(lm2_7,tmp2)]
#head(tmp)
ggplot(tmp2[Date=='2019-05-08'] ,aes(x=datetime)) +
geom_line(aes(y=actual,color='real')) +
geom_line(aes(y=predicted,color='predicted'))
tmp2[,residual:=actual-predicted]
fit_res_tree=rpart(residual~
loc1_DSWRF_surface+loc1_RH_2.m.above.ground+loc1_TCDC_low.cloud.layer+loc1_TMP_2.m.above.ground+loc1_ws_10m+
loc2_DSWRF_surface+loc2_RH_2.m.above.ground+loc2_TCDC_low.cloud.layer+loc2_TMP_2.m.above.ground+loc2_wdir_10m+loc2_ws_10m+
loc3_DSWRF_surface+loc3_RH_2.m.above.ground+loc3_ws_10m+
loc4_DSWRF_surface+loc4_TCDC_low.cloud.layer+loc4_ws_10m+
loc5_DSWRF_surface+loc5_RH_2.m.above.ground+loc5_TCDC_low.cloud.layer+loc5_wdir_10m+loc5_ws_10m+
loc6_DSWRF_surface+loc6_RH_2.m.above.ground+loc6_TCDC_low.cloud.layer+loc6_wdir_10m+loc6_ws_10m+
loc7_DSWRF_surface+loc7_RH_2.m.above.ground+loc7_TCDC_low.cloud.layer,tmp2,
control=rpart.control(cp=0,maxdepth=3))
fancyRpartPlot(fit_res_tree)
tmp2[,loc4ws:=as.numeric(loc4_ws_10m<3.5)]
tmp2[,loc6rh:=as.numeric(loc6_RH_2.m.above.ground<40)]
tmp2[,loc1tmp2ag:=as.numeric(loc1_TMP_2.m.above.ground>=15)]
lm2_8=lm(net~trnd+w_day+mon+hour+prevday+prevweek+loc7tmp2ag:pv:loc7tmp2ag2+loc5tmp2ag:pd:workday+loc4tmp2ag:whour:loc3wdir+loc6tmp2ag:loc7ws:loc1wdir
+loc4wdir:loc7wdir+loc3tmp2ag+loc3tcdc:mon1:loc4rh+loc4ws:loc6rh:loc1tmp2ag,
tmp2)
summary(lm2_8)##
## Call:
## lm(formula = net ~ trnd + w_day + mon + hour + prevday + prevweek +
## loc7tmp2ag:pv:loc7tmp2ag2 + loc5tmp2ag:pd:workday + loc4tmp2ag:whour:loc3wdir +
## loc6tmp2ag:loc7ws:loc1wdir + loc4wdir:loc7wdir + loc3tmp2ag +
## loc3tcdc:mon1:loc4rh + loc4ws:loc6rh:loc1tmp2ag, data = tmp2)
##
## Residuals:
## Min 1Q Median 3Q Max
## -4935.7 -506.7 -2.3 490.3 4654.7
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) -3.579e+02 3.628e+01 -9.866 < 2e-16 ***
## trnd 1.422e-02 7.244e-04 19.624 < 2e-16 ***
## w_dayCum 5.407e+01 1.945e+01 2.779 0.005449 **
## w_dayÇar 9.054e+01 1.946e+01 4.653 3.28e-06 ***
## w_dayPaz -1.543e+02 1.950e+01 -7.912 2.62e-15 ***
## w_dayPer 1.376e+02 1.947e+01 7.070 1.58e-12 ***
## w_dayPts 2.133e+02 1.957e+01 10.898 < 2e-16 ***
## w_daySal 1.526e+02 1.947e+01 7.837 4.79e-15 ***
## monAra 8.991e+01 2.894e+01 3.106 0.001897 **
## monEki 7.890e+01 2.726e+01 2.894 0.003805 **
## monEyl -8.390e+01 2.632e+01 -3.188 0.001434 **
## monHaz 1.208e+02 2.638e+01 4.579 4.69e-06 ***
## monKas 1.124e+02 2.866e+01 3.920 8.86e-05 ***
## monMar 2.759e+01 2.916e+01 0.946 0.344030
## monMay -4.724e+01 2.669e+01 -1.770 0.076703 .
## monNis -8.934e+01 2.877e+01 -3.105 0.001903 **
## monOca 6.580e+01 2.853e+01 2.306 0.021111 *
## monŞub 1.077e+02 3.003e+01 3.586 0.000336 ***
## monTem 8.342e+01 2.561e+01 3.257 0.001125 **
## hour1 1.047e+02 3.598e+01 2.910 0.003615 **
## hour2 7.740e+01 3.598e+01 2.151 0.031495 *
## hour3 5.511e+01 3.600e+01 1.531 0.125829
## hour4 3.068e+01 3.601e+01 0.852 0.394279
## hour5 2.499e+01 3.604e+01 0.693 0.488073
## hour6 -3.602e+01 3.606e+01 -0.999 0.317875
## hour7 -2.749e+01 3.603e+01 -0.763 0.445465
## hour8 6.110e+01 3.616e+01 1.689 0.091146 .
## hour9 1.673e+02 3.630e+01 4.609 4.06e-06 ***
## hour10 1.008e+02 3.641e+01 2.768 0.005636 **
## hour11 1.153e+02 3.662e+01 3.150 0.001635 **
## hour12 -4.214e+01 3.669e+01 -1.148 0.250791
## hour13 -9.858e+01 3.676e+01 -2.682 0.007327 **
## hour14 4.826e+01 3.688e+01 1.308 0.190768
## hour15 4.126e+01 3.694e+01 1.117 0.263973
## hour16 1.883e+01 3.687e+01 0.511 0.609622
## hour17 7.158e+01 3.692e+01 1.939 0.052513 .
## hour18 1.005e+02 3.662e+01 2.744 0.006070 **
## hour19 9.730e+01 3.643e+01 2.671 0.007568 **
## hour20 1.208e+02 3.626e+01 3.331 0.000867 ***
## hour21 1.590e+02 3.618e+01 4.395 1.11e-05 ***
## hour22 1.340e+02 3.605e+01 3.717 0.000202 ***
## hour23 1.144e+02 3.600e+01 3.178 0.001483 **
## prevday 3.290e-01 5.584e-03 58.911 < 2e-16 ***
## prevweek 6.705e-02 5.279e-03 12.701 < 2e-16 ***
## loc3tmp2ag 2.267e+02 1.868e+01 12.138 < 2e-16 ***
## loc4wdir:loc7wdir 2.680e+02 1.709e+01 15.683 < 2e-16 ***
## loc7tmp2ag:pv:loc7tmp2ag2 3.757e+03 1.593e+02 23.585 < 2e-16 ***
## loc5tmp2ag:pd:workday 1.658e+03 7.341e+01 22.580 < 2e-16 ***
## loc4tmp2ag:whour:loc3wdir 8.935e+01 2.038e+01 4.383 1.17e-05 ***
## loc6tmp2ag:loc7ws:loc1wdir 1.366e+03 9.413e+01 14.510 < 2e-16 ***
## loc3tcdc:mon1:loc4rh 4.415e+02 1.909e+01 23.124 < 2e-16 ***
## loc4ws:loc6rh:loc1tmp2ag 3.255e+02 2.487e+01 13.084 < 2e-16 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 849.8 on 26756 degrees of freedom
## Multiple R-squared: 0.3472, Adjusted R-squared: 0.3459
## F-statistic: 279 on 51 and 26756 DF, p-value: < 2.2e-16tmp2[,predicted:=predict(lm2_8,tmp2)]
#head(tmp)
ggplot(tmp2[Date=='2019-05-08'] ,aes(x=datetime)) +
geom_line(aes(y=actual,color='real')) +
geom_line(aes(y=predicted,color='predicted'))
tmp2[,residual:=actual-predicted]
fit_res_tree=rpart(residual~
loc1_DSWRF_surface+loc1_RH_2.m.above.ground+loc1_TCDC_low.cloud.layer+loc1_ws_10m+
loc2_DSWRF_surface+loc2_RH_2.m.above.ground+loc2_TCDC_low.cloud.layer+loc2_TMP_2.m.above.ground+loc2_wdir_10m+loc2_ws_10m+
loc3_DSWRF_surface+loc3_RH_2.m.above.ground+loc3_ws_10m+
loc4_DSWRF_surface+loc4_TCDC_low.cloud.layer+
loc5_DSWRF_surface+loc5_RH_2.m.above.ground+loc5_TCDC_low.cloud.layer+loc5_wdir_10m+loc5_ws_10m+
loc6_DSWRF_surface+loc6_TCDC_low.cloud.layer+loc6_wdir_10m+loc6_ws_10m+
loc7_DSWRF_surface+loc7_RH_2.m.above.ground+loc7_TCDC_low.cloud.layer,tmp2,
control=rpart.control(cp=0,maxdepth=3))
fancyRpartPlot(fit_res_tree)
tmp2[,loc5ws:=as.numeric(loc5_ws_10m<2.4)]
tmp2[,loc2tmp2ag:=as.numeric(loc2_TMP_2.m.above.ground<2.3)]
tmp2[,loc7rh2:=as.numeric(loc7_RH_2.m.above.ground<63)]
lm2_9=lm(net~trnd+w_day+mon+hour+prevday+prevweek+loc7tmp2ag:pv:loc7tmp2ag2+loc5tmp2ag:pd:workday+loc4tmp2ag:whour:loc3wdir+loc6tmp2ag:loc7ws:loc1wdir
+loc4wdir:loc7wdir+loc3tmp2ag+loc3tcdc:mon1:loc4rh+loc4ws:loc6rh:loc1tmp2ag+loc5ws:loc2tmp2ag:loc7rh2,
tmp2)
summary(lm2_9)##
## Call:
## lm(formula = net ~ trnd + w_day + mon + hour + prevday + prevweek +
## loc7tmp2ag:pv:loc7tmp2ag2 + loc5tmp2ag:pd:workday + loc4tmp2ag:whour:loc3wdir +
## loc6tmp2ag:loc7ws:loc1wdir + loc4wdir:loc7wdir + loc3tmp2ag +
## loc3tcdc:mon1:loc4rh + loc4ws:loc6rh:loc1tmp2ag + loc5ws:loc2tmp2ag:loc7rh2,
## data = tmp2)
##
## Residuals:
## Min 1Q Median 3Q Max
## -4938.5 -506.4 -2.0 491.1 4657.6
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) -3.558e+02 3.626e+01 -9.813 < 2e-16 ***
## trnd 1.401e-02 7.250e-04 19.318 < 2e-16 ***
## w_dayCum 5.363e+01 1.944e+01 2.758 0.005813 **
## w_dayÇar 9.175e+01 1.945e+01 4.717 2.40e-06 ***
## w_dayPaz -1.529e+02 1.949e+01 -7.842 4.62e-15 ***
## w_dayPer 1.377e+02 1.946e+01 7.075 1.53e-12 ***
## w_dayPts 2.131e+02 1.956e+01 10.895 < 2e-16 ***
## w_daySal 1.541e+02 1.946e+01 7.920 2.47e-15 ***
## monAra 8.825e+01 2.893e+01 3.050 0.002287 **
## monEki 7.901e+01 2.725e+01 2.900 0.003739 **
## monEyl -8.387e+01 2.630e+01 -3.189 0.001431 **
## monHaz 1.203e+02 2.637e+01 4.564 5.03e-06 ***
## monKas 1.130e+02 2.864e+01 3.944 8.05e-05 ***
## monMar 2.807e+01 2.914e+01 0.963 0.335428
## monMay -4.805e+01 2.667e+01 -1.801 0.071646 .
## monNis -8.982e+01 2.875e+01 -3.124 0.001789 **
## monOca 5.901e+01 2.854e+01 2.067 0.038718 *
## monŞub 1.073e+02 3.001e+01 3.577 0.000349 ***
## monTem 8.331e+01 2.559e+01 3.255 0.001135 **
## hour1 1.042e+02 3.596e+01 2.898 0.003758 **
## hour2 7.803e+01 3.597e+01 2.170 0.030044 *
## hour3 5.586e+01 3.598e+01 1.553 0.120526
## hour4 3.092e+01 3.599e+01 0.859 0.390276
## hour5 2.535e+01 3.603e+01 0.704 0.481652
## hour6 -3.395e+01 3.604e+01 -0.942 0.346298
## hour7 -2.587e+01 3.601e+01 -0.718 0.472477
## hour8 5.946e+01 3.615e+01 1.645 0.099999 .
## hour9 1.668e+02 3.628e+01 4.598 4.29e-06 ***
## hour10 1.014e+02 3.639e+01 2.786 0.005336 **
## hour11 1.165e+02 3.660e+01 3.182 0.001465 **
## hour12 -4.108e+01 3.667e+01 -1.120 0.262666
## hour13 -9.753e+01 3.674e+01 -2.655 0.007944 **
## hour14 4.932e+01 3.686e+01 1.338 0.180981
## hour15 4.230e+01 3.692e+01 1.146 0.251818
## hour16 1.989e+01 3.685e+01 0.540 0.589384
## hour17 7.265e+01 3.690e+01 1.969 0.048953 *
## hour18 1.017e+02 3.660e+01 2.778 0.005466 **
## hour19 9.883e+01 3.641e+01 2.714 0.006643 **
## hour20 1.220e+02 3.624e+01 3.366 0.000764 ***
## hour21 1.598e+02 3.616e+01 4.419 9.96e-06 ***
## hour22 1.350e+02 3.604e+01 3.745 0.000181 ***
## hour23 1.149e+02 3.598e+01 3.194 0.001403 **
## prevday 3.291e-01 5.581e-03 58.958 < 2e-16 ***
## prevweek 6.689e-02 5.276e-03 12.677 < 2e-16 ***
## loc3tmp2ag 2.263e+02 1.867e+01 12.121 < 2e-16 ***
## loc4wdir:loc7wdir 2.678e+02 1.708e+01 15.680 < 2e-16 ***
## loc7tmp2ag:pv:loc7tmp2ag2 3.758e+03 1.592e+02 23.605 < 2e-16 ***
## loc5tmp2ag:pd:workday 1.658e+03 7.337e+01 22.600 < 2e-16 ***
## loc4tmp2ag:whour:loc3wdir 7.994e+01 2.044e+01 3.910 9.25e-05 ***
## loc6tmp2ag:loc7ws:loc1wdir 1.367e+03 9.408e+01 14.526 < 2e-16 ***
## loc3tcdc:mon1:loc4rh 4.429e+02 1.908e+01 23.208 < 2e-16 ***
## loc4ws:loc6rh:loc1tmp2ag 3.256e+02 2.486e+01 13.096 < 2e-16 ***
## loc5ws:loc2tmp2ag:loc7rh2 6.218e+02 1.139e+02 5.460 4.81e-08 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 849.3 on 26755 degrees of freedom
## Multiple R-squared: 0.3479, Adjusted R-squared: 0.3466
## F-statistic: 274.5 on 52 and 26755 DF, p-value: < 2.2e-16tmp2[,predicted:=predict(lm2_9,tmp2)]
#head(tmp)
ggplot(tmp2[Date=='2019-05-08'] ,aes(x=datetime)) +
geom_line(aes(y=actual,color='real')) +
geom_line(aes(y=predicted,color='predicted'))
tmp2[,residual:=actual-predicted]
fit_res_tree=rpart(residual~
loc1_DSWRF_surface+loc1_RH_2.m.above.ground+loc1_TCDC_low.cloud.layer+loc1_ws_10m+
loc2_DSWRF_surface+loc2_RH_2.m.above.ground+loc2_TCDC_low.cloud.layer+loc2_wdir_10m+loc2_ws_10m+
loc3_DSWRF_surface+loc3_RH_2.m.above.ground+loc3_ws_10m+
loc4_DSWRF_surface+loc4_TCDC_low.cloud.layer+
loc5_DSWRF_surface+loc5_RH_2.m.above.ground+loc5_TCDC_low.cloud.layer+loc5_wdir_10m+
loc6_DSWRF_surface+loc6_TCDC_low.cloud.layer+loc6_wdir_10m+loc6_ws_10m+
loc7_DSWRF_surface+loc7_TCDC_low.cloud.layer,tmp2,
control=rpart.control(cp=0,maxdepth=3))
fancyRpartPlot(fit_res_tree)
tmp2[,loc5rh:=as.numeric(loc5_RH_2.m.above.ground<16)]
tmp2[,loc1ws:=as.numeric(loc1_ws_10m>=3.9)]
tmp2[,loc2dswrf:=as.numeric(loc2_DSWRF_surface<841)]
lm2_10=lm(net~trnd+w_day+mon+hour+prevday+prevweek+loc7tmp2ag:pv:loc7tmp2ag2+loc5tmp2ag:pd:workday+loc4tmp2ag:whour:loc3wdir+loc6tmp2ag:loc7ws:loc1wdir
+loc4wdir:loc7wdir+loc3tmp2ag+loc3tcdc:mon1:loc4rh+loc4ws:loc6rh:loc1tmp2ag+loc5ws:loc2tmp2ag:loc7rh2+loc5rh:loc1ws:loc2dswrf,
tmp2)
summary(lm2_10)##
## Call:
## lm(formula = net ~ trnd + w_day + mon + hour + prevday + prevweek +
## loc7tmp2ag:pv:loc7tmp2ag2 + loc5tmp2ag:pd:workday + loc4tmp2ag:whour:loc3wdir +
## loc6tmp2ag:loc7ws:loc1wdir + loc4wdir:loc7wdir + loc3tmp2ag +
## loc3tcdc:mon1:loc4rh + loc4ws:loc6rh:loc1tmp2ag + loc5ws:loc2tmp2ag:loc7rh2 +
## loc5rh:loc1ws:loc2dswrf, data = tmp2)
##
## Residuals:
## Min 1Q Median 3Q Max
## -4930.8 -504.4 -2.4 490.0 4661.5
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) -3.683e+02 3.621e+01 -10.171 < 2e-16 ***
## trnd 1.406e-02 7.236e-04 19.435 < 2e-16 ***
## w_dayCum 5.274e+01 1.940e+01 2.718 0.006569 **
## w_dayÇar 9.070e+01 1.941e+01 4.673 2.98e-06 ***
## w_dayPaz -1.538e+02 1.946e+01 -7.904 2.80e-15 ***
## w_dayPer 1.353e+02 1.942e+01 6.968 3.28e-12 ***
## w_dayPts 2.107e+02 1.952e+01 10.793 < 2e-16 ***
## w_daySal 1.512e+02 1.942e+01 7.784 7.28e-15 ***
## monAra 1.032e+02 2.891e+01 3.571 0.000357 ***
## monEki 9.234e+01 2.722e+01 3.392 0.000695 ***
## monEyl -7.390e+01 2.627e+01 -2.813 0.004904 **
## monHaz 1.380e+02 2.637e+01 5.233 1.68e-07 ***
## monKas 1.274e+02 2.862e+01 4.452 8.53e-06 ***
## monMar 4.284e+01 2.912e+01 1.471 0.141219
## monMay -3.354e+01 2.666e+01 -1.258 0.208302
## monNis -7.557e+01 2.873e+01 -2.631 0.008529 **
## monOca 7.411e+01 2.852e+01 2.598 0.009370 **
## monŞub 1.225e+02 2.998e+01 4.084 4.44e-05 ***
## monTem 1.041e+02 2.562e+01 4.062 4.87e-05 ***
## hour1 1.044e+02 3.589e+01 2.909 0.003628 **
## hour2 7.807e+01 3.589e+01 2.175 0.029641 *
## hour3 5.581e+01 3.591e+01 1.554 0.120106
## hour4 3.081e+01 3.592e+01 0.858 0.391020
## hour5 2.516e+01 3.595e+01 0.700 0.483975
## hour6 -3.428e+01 3.597e+01 -0.953 0.340672
## hour7 -2.598e+01 3.594e+01 -0.723 0.469811
## hour8 5.921e+01 3.607e+01 1.641 0.100709
## hour9 1.669e+02 3.621e+01 4.609 4.06e-06 ***
## hour10 1.014e+02 3.632e+01 2.792 0.005241 **
## hour11 1.171e+02 3.653e+01 3.205 0.001350 **
## hour12 -4.266e+01 3.660e+01 -1.166 0.243727
## hour13 -1.023e+02 3.667e+01 -2.790 0.005267 **
## hour14 4.503e+01 3.679e+01 1.224 0.220961
## hour15 3.558e+01 3.685e+01 0.966 0.334295
## hour16 7.525e+00 3.680e+01 0.204 0.837981
## hour17 6.698e+01 3.683e+01 1.819 0.068957 .
## hour18 1.011e+02 3.652e+01 2.768 0.005642 **
## hour19 1.015e+02 3.634e+01 2.794 0.005215 **
## hour20 1.231e+02 3.617e+01 3.402 0.000669 ***
## hour21 1.606e+02 3.608e+01 4.452 8.54e-06 ***
## hour22 1.353e+02 3.596e+01 3.763 0.000168 ***
## hour23 1.152e+02 3.591e+01 3.209 0.001332 **
## prevday 3.283e-01 5.571e-03 58.930 < 2e-16 ***
## prevweek 6.614e-02 5.266e-03 12.559 < 2e-16 ***
## loc3tmp2ag 2.211e+02 1.864e+01 11.862 < 2e-16 ***
## loc4wdir:loc7wdir 2.687e+02 1.705e+01 15.765 < 2e-16 ***
## loc7tmp2ag:pv:loc7tmp2ag2 3.575e+03 1.598e+02 22.364 < 2e-16 ***
## loc5tmp2ag:pd:workday 1.508e+03 7.463e+01 20.199 < 2e-16 ***
## loc4tmp2ag:whour:loc3wdir 7.798e+01 2.040e+01 3.822 0.000133 ***
## loc6tmp2ag:loc7ws:loc1wdir 1.391e+03 9.392e+01 14.815 < 2e-16 ***
## loc3tcdc:mon1:loc4rh 4.449e+02 1.905e+01 23.359 < 2e-16 ***
## loc4ws:loc6rh:loc1tmp2ag 3.345e+02 2.483e+01 13.473 < 2e-16 ***
## loc5ws:loc2tmp2ag:loc7rh2 6.214e+02 1.137e+02 5.467 4.61e-08 ***
## loc5rh:loc1ws:loc2dswrf 1.160e+03 1.112e+02 10.435 < 2e-16 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 847.6 on 26754 degrees of freedom
## Multiple R-squared: 0.3505, Adjusted R-squared: 0.3492
## F-statistic: 272.4 on 53 and 26754 DF, p-value: < 2.2e-16tmp2[,predicted:=predict(lm2_10,tmp2)]
#head(tmp)
ggplot(tmp2[Date=='2019-05-08'] ,aes(x=datetime)) +
geom_line(aes(y=actual,color='real')) +
geom_line(aes(y=predicted,color='predicted'))
tmp2[,residual:=actual-predicted]
fit_res_tree=rpart(residual~
loc1_DSWRF_surface+loc1_RH_2.m.above.ground+loc1_TCDC_low.cloud.layer+
loc2_RH_2.m.above.ground+loc2_TCDC_low.cloud.layer+loc2_wdir_10m+loc2_ws_10m+
loc3_DSWRF_surface+loc3_RH_2.m.above.ground+loc3_ws_10m+
loc4_DSWRF_surface+loc4_TCDC_low.cloud.layer+
loc5_DSWRF_surface+loc5_TCDC_low.cloud.layer+loc5_wdir_10m+
loc6_DSWRF_surface+loc6_TCDC_low.cloud.layer+loc6_wdir_10m+loc6_ws_10m+
loc7_DSWRF_surface+loc7_TCDC_low.cloud.layer,tmp2,
control=rpart.control(cp=0,maxdepth=3))
fancyRpartPlot(fit_res_tree)
tmp2[,loc6tcdc:=as.numeric(loc6_TCDC_low.cloud.layer>=84)]
tmp2[,loc1dswrf:=as.numeric(loc1_DSWRF_surface>=90)]
tmp2[,loc2rh2:=as.numeric(loc2_RH_2.m.above.ground>=91)]
lm2_11=lm(net~trnd+w_day+mon+hour+prevday+prevweek+loc7tmp2ag:pv:loc7tmp2ag2+loc5tmp2ag:pd:workday+loc4tmp2ag:whour:loc3wdir+loc6tmp2ag:loc7ws:loc1wdir
+loc4wdir:loc7wdir+loc3tmp2ag+loc3tcdc:mon1:loc4rh+loc4ws:loc6rh:loc1tmp2ag+loc5ws:loc2tmp2ag:loc7rh2+loc5rh:loc1ws:loc2dswrf
+loc6tcdc:loc1dswrf:loc2rh2,
tmp2)
summary(lm2_11)##
## Call:
## lm(formula = net ~ trnd + w_day + mon + hour + prevday + prevweek +
## loc7tmp2ag:pv:loc7tmp2ag2 + loc5tmp2ag:pd:workday + loc4tmp2ag:whour:loc3wdir +
## loc6tmp2ag:loc7ws:loc1wdir + loc4wdir:loc7wdir + loc3tmp2ag +
## loc3tcdc:mon1:loc4rh + loc4ws:loc6rh:loc1tmp2ag + loc5ws:loc2tmp2ag:loc7rh2 +
## loc5rh:loc1ws:loc2dswrf + loc6tcdc:loc1dswrf:loc2rh2, data = tmp2)
##
## Residuals:
## Min 1Q Median 3Q Max
## -4927.5 -505.1 -1.9 490.9 4594.2
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) -3.671e+02 3.614e+01 -10.158 < 2e-16 ***
## trnd 1.404e-02 7.221e-04 19.437 < 2e-16 ***
## w_dayCum 5.393e+01 1.937e+01 2.785 0.005359 **
## w_dayÇar 9.022e+01 1.937e+01 4.658 3.21e-06 ***
## w_dayPaz -1.543e+02 1.942e+01 -7.949 1.96e-15 ***
## w_dayPer 1.361e+02 1.938e+01 7.024 2.20e-12 ***
## w_dayPts 2.067e+02 1.949e+01 10.607 < 2e-16 ***
## w_daySal 1.508e+02 1.938e+01 7.781 7.43e-15 ***
## monAra 9.568e+01 2.886e+01 3.315 0.000916 ***
## monEki 9.305e+01 2.717e+01 3.425 0.000616 ***
## monEyl -7.362e+01 2.621e+01 -2.809 0.004979 **
## monHaz 1.381e+02 2.631e+01 5.249 1.54e-07 ***
## monKas 1.297e+02 2.856e+01 4.542 5.60e-06 ***
## monMar 4.380e+01 2.906e+01 1.507 0.131723
## monMay -3.203e+01 2.660e+01 -1.204 0.228620
## monNis -7.300e+01 2.867e+01 -2.546 0.010904 *
## monOca 7.250e+01 2.846e+01 2.547 0.010874 *
## monŞub 1.193e+02 2.993e+01 3.985 6.76e-05 ***
## monTem 1.039e+02 2.557e+01 4.064 4.83e-05 ***
## hour1 1.046e+02 3.582e+01 2.921 0.003492 **
## hour2 7.824e+01 3.582e+01 2.184 0.028950 *
## hour3 5.590e+01 3.583e+01 1.560 0.118808
## hour4 3.076e+01 3.585e+01 0.858 0.390833
## hour5 2.513e+01 3.588e+01 0.700 0.483687
## hour6 -3.438e+01 3.590e+01 -0.958 0.338255
## hour7 -2.612e+01 3.587e+01 -0.728 0.466526
## hour8 5.942e+01 3.600e+01 1.651 0.098851 .
## hour9 1.675e+02 3.613e+01 4.635 3.59e-06 ***
## hour10 9.775e+01 3.624e+01 2.697 0.007001 **
## hour11 1.153e+02 3.645e+01 3.164 0.001558 **
## hour12 -4.435e+01 3.652e+01 -1.214 0.224615
## hour13 -1.054e+02 3.660e+01 -2.881 0.003966 **
## hour14 3.938e+01 3.672e+01 1.072 0.283549
## hour15 3.005e+01 3.678e+01 0.817 0.413907
## hour16 4.650e+00 3.673e+01 0.127 0.899238
## hour17 6.705e+01 3.675e+01 1.824 0.068101 .
## hour18 9.807e+01 3.645e+01 2.690 0.007144 **
## hour19 1.011e+02 3.626e+01 2.788 0.005304 **
## hour20 1.211e+02 3.609e+01 3.355 0.000796 ***
## hour21 1.600e+02 3.601e+01 4.444 8.86e-06 ***
## hour22 1.356e+02 3.589e+01 3.778 0.000158 ***
## hour23 1.154e+02 3.583e+01 3.222 0.001277 **
## prevday 3.278e-01 5.560e-03 58.960 < 2e-16 ***
## prevweek 6.562e-02 5.256e-03 12.485 < 2e-16 ***
## loc3tmp2ag 2.230e+02 1.860e+01 11.988 < 2e-16 ***
## loc4wdir:loc7wdir 2.696e+02 1.701e+01 15.850 < 2e-16 ***
## loc7tmp2ag:pv:loc7tmp2ag2 3.578e+03 1.595e+02 22.428 < 2e-16 ***
## loc5tmp2ag:pd:workday 1.509e+03 7.448e+01 20.262 < 2e-16 ***
## loc4tmp2ag:whour:loc3wdir 7.891e+01 2.036e+01 3.875 0.000107 ***
## loc6tmp2ag:loc7ws:loc1wdir 1.393e+03 9.373e+01 14.866 < 2e-16 ***
## loc3tcdc:mon1:loc4rh 4.350e+02 1.903e+01 22.855 < 2e-16 ***
## loc4ws:loc6rh:loc1tmp2ag 3.342e+02 2.478e+01 13.487 < 2e-16 ***
## loc5ws:loc2tmp2ag:loc7rh2 6.243e+02 1.134e+02 5.504 3.74e-08 ***
## loc5rh:loc1ws:loc2dswrf 1.162e+03 1.110e+02 10.468 < 2e-16 ***
## loc6tcdc:loc1dswrf:loc2rh2 1.595e+03 1.527e+02 10.444 < 2e-16 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 845.9 on 26753 degrees of freedom
## Multiple R-squared: 0.3532, Adjusted R-squared: 0.3519
## F-statistic: 270.5 on 54 and 26753 DF, p-value: < 2.2e-16tmp2[,predicted:=predict(lm2_11,tmp2)]
#head(tmp)
ggplot(tmp2[Date=='2019-05-08'] ,aes(x=datetime)) +
geom_line(aes(y=actual,color='real')) +
geom_line(aes(y=predicted,color='predicted'))
tmp2[,residual:=actual-predicted]
fit_res_tree=rpart(residual~
loc1_RH_2.m.above.ground+loc1_TCDC_low.cloud.layer+
loc2_TCDC_low.cloud.layer+loc2_wdir_10m+loc2_ws_10m+
loc3_DSWRF_surface+loc3_RH_2.m.above.ground+loc3_ws_10m+
loc4_DSWRF_surface+loc4_TCDC_low.cloud.layer+
loc5_DSWRF_surface+loc5_TCDC_low.cloud.layer+loc5_wdir_10m+
loc6_DSWRF_surface+loc6_wdir_10m+loc6_ws_10m+
loc7_DSWRF_surface+loc7_TCDC_low.cloud.layer,tmp2,
control=rpart.control(cp=0,maxdepth=3))
fancyRpartPlot(fit_res_tree)
tmp2[,loc3ws:=as.numeric(loc3_ws_10m<1.1)]
tmp2[,loc1tcdc:=as.numeric(loc1_TCDC_low.cloud.layer>=63)]
tmp2[,loc5wdir:=as.numeric(loc5_wdir_10m<69)]
lm2_12=lm(net~trnd+w_day+mon+hour+prevday+prevweek+loc7tmp2ag:pv:loc7tmp2ag2+loc5tmp2ag:pd:workday+loc4tmp2ag:whour:loc3wdir+loc6tmp2ag:loc7ws:loc1wdir
+loc4wdir:loc7wdir+loc3tmp2ag+loc3tcdc:mon1:loc4rh+loc4ws:loc6rh:loc1tmp2ag+loc5ws:loc2tmp2ag:loc7rh2+loc5rh:loc1ws:loc2dswrf
+loc6tcdc:loc1dswrf:loc2rh2+loc3ws:loc1tcdc:loc5wdir,
tmp2)
summary(lm2_12)##
## Call:
## lm(formula = net ~ trnd + w_day + mon + hour + prevday + prevweek +
## loc7tmp2ag:pv:loc7tmp2ag2 + loc5tmp2ag:pd:workday + loc4tmp2ag:whour:loc3wdir +
## loc6tmp2ag:loc7ws:loc1wdir + loc4wdir:loc7wdir + loc3tmp2ag +
## loc3tcdc:mon1:loc4rh + loc4ws:loc6rh:loc1tmp2ag + loc5ws:loc2tmp2ag:loc7rh2 +
## loc5rh:loc1ws:loc2dswrf + loc6tcdc:loc1dswrf:loc2rh2 + loc3ws:loc1tcdc:loc5wdir,
## data = tmp2)
##
## Residuals:
## Min 1Q Median 3Q Max
## -4931.9 -504.3 -1.2 492.1 4599.4
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) -3.658e+02 3.609e+01 -10.136 < 2e-16 ***
## trnd 1.389e-02 7.214e-04 19.260 < 2e-16 ***
## w_dayCum 5.387e+01 1.934e+01 2.786 0.005345 **
## w_dayÇar 9.097e+01 1.934e+01 4.702 2.58e-06 ***
## w_dayPaz -1.536e+02 1.939e+01 -7.919 2.48e-15 ***
## w_dayPer 1.343e+02 1.936e+01 6.937 4.10e-12 ***
## w_dayPts 2.054e+02 1.946e+01 10.555 < 2e-16 ***
## w_daySal 1.518e+02 1.936e+01 7.842 4.60e-15 ***
## monAra 9.180e+01 2.882e+01 3.185 0.001450 **
## monEki 9.345e+01 2.713e+01 3.444 0.000574 ***
## monEyl -7.342e+01 2.618e+01 -2.805 0.005042 **
## monHaz 1.381e+02 2.628e+01 5.253 1.51e-07 ***
## monKas 1.286e+02 2.852e+01 4.510 6.52e-06 ***
## monMar 4.000e+01 2.902e+01 1.378 0.168135
## monMay -3.251e+01 2.657e+01 -1.224 0.221043
## monNis -7.400e+01 2.863e+01 -2.584 0.009764 **
## monOca 6.647e+01 2.844e+01 2.337 0.019426 *
## monŞub 1.180e+02 2.989e+01 3.949 7.86e-05 ***
## monTem 1.039e+02 2.553e+01 4.068 4.75e-05 ***
## hour1 1.056e+02 3.577e+01 2.953 0.003151 **
## hour2 7.717e+01 3.577e+01 2.157 0.030989 *
## hour3 5.794e+01 3.579e+01 1.619 0.105458
## hour4 3.175e+01 3.580e+01 0.887 0.375113
## hour5 2.610e+01 3.583e+01 0.728 0.466468
## hour6 -3.233e+01 3.585e+01 -0.902 0.367117
## hour7 -2.515e+01 3.582e+01 -0.702 0.482564
## hour8 5.849e+01 3.595e+01 1.627 0.103795
## hour9 1.665e+02 3.609e+01 4.614 3.96e-06 ***
## hour10 9.722e+01 3.620e+01 2.686 0.007237 **
## hour11 1.157e+02 3.641e+01 3.179 0.001479 **
## hour12 -4.388e+01 3.648e+01 -1.203 0.229001
## hour13 -1.037e+02 3.655e+01 -2.839 0.004534 **
## hour14 4.131e+01 3.667e+01 1.126 0.260039
## hour15 3.201e+01 3.673e+01 0.872 0.383481
## hour16 5.314e+00 3.668e+01 0.145 0.884810
## hour17 6.746e+01 3.670e+01 1.838 0.066094 .
## hour18 9.763e+01 3.640e+01 2.682 0.007326 **
## hour19 1.004e+02 3.622e+01 2.772 0.005579 **
## hour20 1.204e+02 3.605e+01 3.341 0.000834 ***
## hour21 1.613e+02 3.596e+01 4.485 7.32e-06 ***
## hour22 1.377e+02 3.584e+01 3.842 0.000122 ***
## hour23 1.175e+02 3.579e+01 3.284 0.001026 **
## prevday 3.276e-01 5.552e-03 59.000 < 2e-16 ***
## prevweek 6.601e-02 5.249e-03 12.575 < 2e-16 ***
## loc3tmp2ag 2.231e+02 1.858e+01 12.012 < 2e-16 ***
## loc4wdir:loc7wdir 2.689e+02 1.699e+01 15.827 < 2e-16 ***
## loc7tmp2ag:pv:loc7tmp2ag2 3.579e+03 1.593e+02 22.464 < 2e-16 ***
## loc5tmp2ag:pd:workday 1.510e+03 7.438e+01 20.305 < 2e-16 ***
## loc4tmp2ag:whour:loc3wdir 7.994e+01 2.034e+01 3.931 8.49e-05 ***
## loc6tmp2ag:loc7ws:loc1wdir 1.394e+03 9.361e+01 14.889 < 2e-16 ***
## loc3tcdc:mon1:loc4rh 4.340e+02 1.901e+01 22.835 < 2e-16 ***
## loc4ws:loc6rh:loc1tmp2ag 3.346e+02 2.474e+01 13.522 < 2e-16 ***
## loc5ws:loc2tmp2ag:loc7rh2 6.304e+02 1.133e+02 5.564 2.65e-08 ***
## loc5rh:loc1ws:loc2dswrf 1.161e+03 1.108e+02 10.478 < 2e-16 ***
## loc6tcdc:loc1dswrf:loc2rh2 1.448e+03 1.534e+02 9.441 < 2e-16 ***
## loc3ws:loc1tcdc:loc5wdir 1.165e+03 1.365e+02 8.534 < 2e-16 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 844.8 on 26752 degrees of freedom
## Multiple R-squared: 0.3549, Adjusted R-squared: 0.3536
## F-statistic: 267.6 on 55 and 26752 DF, p-value: < 2.2e-16tmp2[,predicted:=predict(lm2_12,tmp2)]
#head(tmp)
ggplot(tmp2[Date=='2019-05-08'] ,aes(x=datetime)) +
geom_line(aes(y=actual,color='real')) +
geom_line(aes(y=predicted,color='predicted'))
tmp2[,residual:=actual-predicted]
fit_res_tree=rpart(residual~
loc1_RH_2.m.above.ground+
loc2_TCDC_low.cloud.layer+loc2_wdir_10m+loc2_ws_10m+
loc3_DSWRF_surface+loc3_RH_2.m.above.ground+
loc4_DSWRF_surface+loc4_TCDC_low.cloud.layer+
loc5_DSWRF_surface+loc5_TCDC_low.cloud.layer+
loc6_DSWRF_surface+loc6_wdir_10m+loc6_ws_10m+
loc7_DSWRF_surface+loc7_TCDC_low.cloud.layer,tmp2,
control=rpart.control(cp=0,maxdepth=3))
fancyRpartPlot(fit_res_tree)
After comparision between models, deciding to continue with the 13th model
Submit results depends on this model
tmp2[,loc3rh:=as.numeric(loc3_RH_2.m.above.ground<25)]
tmp2[,loc5tcdc:=as.numeric(loc5_TCDC_low.cloud.layer<0.15)]
tmp2[,loc1rh:=as.numeric(loc1_RH_2.m.above.ground>=15)]
lm2_13=lm(net~trnd+w_day+mon+hour+prevday+prevweek+loc7tmp2ag:pv:loc7tmp2ag2+loc5tmp2ag:pd:workday+loc4tmp2ag:whour:loc3wdir+loc6tmp2ag:loc7ws:loc1wdir
+loc4wdir:loc7wdir+loc3tmp2ag+loc3tcdc:mon1:loc4rh+loc4ws:loc6rh:loc1tmp2ag+loc5ws:loc2tmp2ag:loc7rh2+loc5rh:loc1ws:loc2dswrf
+loc6tcdc:loc1dswrf:loc2rh2+loc3ws:loc1tcdc:loc5wdir+loc3rh:loc5tcdc:loc1rh,
tmp2)
summary(lm2_13)##
## Call:
## lm(formula = net ~ trnd + w_day + mon + hour + prevday + prevweek +
## loc7tmp2ag:pv:loc7tmp2ag2 + loc5tmp2ag:pd:workday + loc4tmp2ag:whour:loc3wdir +
## loc6tmp2ag:loc7ws:loc1wdir + loc4wdir:loc7wdir + loc3tmp2ag +
## loc3tcdc:mon1:loc4rh + loc4ws:loc6rh:loc1tmp2ag + loc5ws:loc2tmp2ag:loc7rh2 +
## loc5rh:loc1ws:loc2dswrf + loc6tcdc:loc1dswrf:loc2rh2 + loc3ws:loc1tcdc:loc5wdir +
## loc3rh:loc5tcdc:loc1rh, data = tmp2)
##
## Residuals:
## Min 1Q Median 3Q Max
## -4882.9 -505.1 -1.5 493.3 4688.8
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) -3.727e+02 3.601e+01 -10.349 < 2e-16 ***
## trnd 1.389e-02 7.196e-04 19.306 < 2e-16 ***
## w_dayCum 5.607e+01 1.929e+01 2.906 0.003666 **
## w_dayÇar 8.613e+01 1.930e+01 4.462 8.16e-06 ***
## w_dayPaz -1.550e+02 1.934e+01 -8.015 1.15e-15 ***
## w_dayPer 1.335e+02 1.931e+01 6.912 4.88e-12 ***
## w_dayPts 1.998e+02 1.942e+01 10.290 < 2e-16 ***
## w_daySal 1.480e+02 1.931e+01 7.662 1.90e-14 ***
## monAra 1.095e+02 2.880e+01 3.802 0.000144 ***
## monEki 1.007e+02 2.707e+01 3.720 0.000200 ***
## monEyl -5.841e+01 2.615e+01 -2.234 0.025496 *
## monHaz 1.678e+02 2.635e+01 6.368 1.95e-10 ***
## monKas 1.438e+02 2.849e+01 5.049 4.46e-07 ***
## monMar 5.443e+01 2.898e+01 1.878 0.060394 .
## monMay -3.177e+01 2.650e+01 -1.199 0.230651
## monNis -6.222e+01 2.858e+01 -2.177 0.029509 *
## monOca 8.441e+01 2.841e+01 2.971 0.002970 **
## monŞub 1.345e+02 2.985e+01 4.506 6.64e-06 ***
## monTem 1.189e+02 2.551e+01 4.661 3.17e-06 ***
## hour1 1.056e+02 3.569e+01 2.960 0.003082 **
## hour2 7.598e+01 3.569e+01 2.129 0.033263 *
## hour3 5.697e+01 3.570e+01 1.596 0.110584
## hour4 3.044e+01 3.572e+01 0.852 0.394106
## hour5 2.401e+01 3.575e+01 0.672 0.501753
## hour6 -3.485e+01 3.577e+01 -0.974 0.329837
## hour7 -2.644e+01 3.574e+01 -0.740 0.459403
## hour8 5.715e+01 3.587e+01 1.593 0.111072
## hour9 1.673e+02 3.600e+01 4.647 3.38e-06 ***
## hour10 9.874e+01 3.611e+01 2.734 0.006255 **
## hour11 1.154e+02 3.632e+01 3.177 0.001489 **
## hour12 -4.996e+01 3.639e+01 -1.373 0.169805
## hour13 -1.227e+02 3.650e+01 -3.360 0.000779 ***
## hour14 1.547e+01 3.666e+01 0.422 0.673070
## hour15 2.584e+00 3.673e+01 0.070 0.943925
## hour16 -2.736e+01 3.670e+01 -0.746 0.455937
## hour17 3.577e+01 3.672e+01 0.974 0.329968
## hour18 7.311e+01 3.638e+01 2.009 0.044498 *
## hour19 8.868e+01 3.614e+01 2.454 0.014154 *
## hour20 1.166e+02 3.596e+01 3.242 0.001190 **
## hour21 1.619e+02 3.588e+01 4.513 6.41e-06 ***
## hour22 1.387e+02 3.576e+01 3.878 0.000105 ***
## hour23 1.189e+02 3.570e+01 3.330 0.000869 ***
## prevday 3.246e-01 5.545e-03 58.548 < 2e-16 ***
## prevweek 6.674e-02 5.237e-03 12.745 < 2e-16 ***
## loc3tmp2ag 1.786e+02 1.894e+01 9.426 < 2e-16 ***
## loc4wdir:loc7wdir 2.534e+02 1.700e+01 14.901 < 2e-16 ***
## loc7tmp2ag:pv:loc7tmp2ag2 3.545e+03 1.589e+02 22.304 < 2e-16 ***
## loc5tmp2ag:pd:workday 1.385e+03 7.502e+01 18.467 < 2e-16 ***
## loc4tmp2ag:whour:loc3wdir 7.301e+01 2.030e+01 3.597 0.000322 ***
## loc6tmp2ag:loc7ws:loc1wdir 1.367e+03 9.341e+01 14.638 < 2e-16 ***
## loc3tcdc:mon1:loc4rh 4.477e+02 1.900e+01 23.563 < 2e-16 ***
## loc4ws:loc6rh:loc1tmp2ag 3.097e+02 2.478e+01 12.496 < 2e-16 ***
## loc5ws:loc2tmp2ag:loc7rh2 6.280e+02 1.130e+02 5.557 2.77e-08 ***
## loc5rh:loc1ws:loc2dswrf 1.098e+03 1.107e+02 9.921 < 2e-16 ***
## loc6tcdc:loc1dswrf:loc2rh2 1.453e+03 1.531e+02 9.495 < 2e-16 ***
## loc3ws:loc1tcdc:loc5wdir 1.166e+03 1.362e+02 8.561 < 2e-16 ***
## loc3rh:loc5tcdc:loc1rh 2.835e+02 2.494e+01 11.368 < 2e-16 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 842.8 on 26751 degrees of freedom
## Multiple R-squared: 0.358, Adjusted R-squared: 0.3567
## F-statistic: 266.4 on 56 and 26751 DF, p-value: < 2.2e-16tmp2[,predicted:=predict(lm2_13,tmp2)]checkresiduals(lm2_13$residuals)## Warning in modeldf.default(object): Could not find appropriate degrees of
## freedom for this model.
tmp2[,residuals13 := residuals(lm2_13)]
acf(residuals(lm2_13) , lag.max = 160)
ggplot(tmp2[Date=='2022-01-07'] ,aes(x=datetime)) +
geom_line(aes(y=actual,color='real')) +
geom_line(aes(y=predicted,color='predicted'))
tmp2 = tmp2 %>%
mutate(predicted_diretion =
case_when(predicted < -50 ~ "Negative" ,
predicted>=-50 & predicted<=50 ~ "Neutral" ,
predicted> 50 ~ "Positive" )
)
tmp2= as.data.table(tmp2)
tmp2## datetime net system_direction loc1_DSWRF_surface
## 1: 2019-01-01 00:00:00 -1934.795 Negative 0.00
## 2: 2019-01-01 01:00:00 -750.866 Negative 0.00
## 3: 2019-01-01 02:00:00 -982.767 Negative 0.00
## 4: 2019-01-01 03:00:00 -764.620 Negative 0.00
## 5: 2019-01-01 04:00:00 -1233.289 Negative 0.00
## ---
## 26804: 2022-01-21 19:00:00 0.000 55.38
## 26805: 2022-01-21 20:00:00 0.000 44.30
## 26806: 2022-01-21 21:00:00 0.000 36.92
## 26807: 2022-01-21 22:00:00 0.000 0.00
## 26808: 2022-01-21 23:00:00 0.000 0.00
## loc1_RH_2.m.above.ground loc1_TCDC_low.cloud.layer
## 1: 84.20000 0.00000
## 2: 84.43333 3.00000
## 3: 84.66667 6.00000
## 4: 84.90000 9.00000
## 5: 87.26667 29.66667
## ---
## 26804: 97.90000 99.90000
## 26805: 98.00000 99.90000
## 26806: 98.70000 99.90000
## 26807: 98.80000 100.00000
## 26808: 98.90000 100.00000
## loc1_TMP_2.m.above.ground loc1_wdir_10m loc1_ws_10m loc2_DSWRF_surface
## 1: -4.081 244.986564 2.302671 0.00
## 2: -3.707 245.382871 2.443606 0.00
## 3: -3.333 245.779178 2.584541 0.00
## 4: -2.959 246.175485 2.725476 0.00
## 5: -2.591 240.651199 2.755832 0.00
## ---
## 26804: -1.623 2.212407 1.948613 40.46
## 26805: -1.285 357.806092 1.767906 32.36
## 26806: -0.937 354.741987 2.657864 26.98
## 26807: -0.641 7.030764 3.098056 0.00
## 26808: -0.540 27.723707 3.520917 0.00
## loc2_RH_2.m.above.ground loc2_TCDC_low.cloud.layer
## 1: 59.80000 0.0
## 2: 56.93333 0.0
## 3: 54.06667 0.0
## 4: 51.20000 0.0
## 5: 51.56667 0.0
## ---
## 26804: 66.80000 99.9
## 26805: 68.60000 99.9
## 26806: 68.50000 99.9
## 26807: 73.90000 100.0
## 26808: 79.80000 100.0
## loc2_TMP_2.m.above.ground loc2_wdir_10m loc2_ws_10m loc3_DSWRF_surface
## 1: 3.369000 192.5959 3.057220 0.00
## 2: 3.333000 192.2694 3.202298 0.00
## 3: 3.297000 191.9429 3.347376 0.00
## 4: 3.261000 191.6164 3.492454 0.00
## 5: 3.365667 191.7771 3.608525 0.00
## ---
## 26804: 4.347000 208.0302 3.265933 163.96
## 26805: 4.295000 203.6774 3.355885 131.16
## 26806: 4.533000 206.7301 3.698566 109.30
## 26807: 4.649000 211.2813 4.276995 0.00
## 26808: 4.530000 212.4927 3.987479 0.00
## loc3_RH_2.m.above.ground loc3_TCDC_low.cloud.layer
## 1: 84.40000 1.0000000
## 2: 85.06667 0.6666667
## 3: 85.73333 0.3333333
## 4: 86.40000 0.0000000
## 5: 87.83333 1.3333333
## ---
## 26804: 72.40000 0.0000000
## 26805: 71.90000 0.0000000
## 26806: 72.70000 0.0000000
## 26807: 74.60000 5.0000000
## 26808: 77.70000 4.8000000
## loc3_TMP_2.m.above.ground loc3_wdir_10m loc3_ws_10m loc4_DSWRF_surface
## 1: -6.211000 351.73126 1.784915 0.00
## 2: -5.950333 350.11411 1.928927 0.00
## 3: -5.689667 348.49697 2.072939 0.00
## 4: -5.429000 346.87983 2.216951 0.00
## 5: -5.094333 337.72332 1.997965 0.00
## ---
## 26804: -2.723000 43.76432 3.014659 30.56
## 26805: -3.055000 45.90085 2.955315 24.46
## 26806: -3.107000 42.20943 2.763192 20.36
## 26807: -3.091000 39.53944 2.716282 0.00
## 26808: -2.910000 51.42887 2.849656 0.00
## loc4_RH_2.m.above.ground loc4_TCDC_low.cloud.layer
## 1: 52.20000 0.00000
## 2: 53.90000 9.00000
## 3: 55.60000 18.00000
## 4: 57.30000 27.00000
## 5: 59.76667 51.33333
## ---
## 26804: 88.20000 98.10000
## 26805: 89.20000 98.50000
## 26806: 89.20000 98.70000
## 26807: 89.30000 100.00000
## 26808: 90.10000 100.00000
## loc4_TMP_2.m.above.ground loc4_wdir_10m loc4_ws_10m loc5_DSWRF_surface
## 1: 7.309000 248.9115 5.151732 0.00
## 2: 7.203000 248.8854 5.512422 0.00
## 3: 7.097000 248.8594 5.873112 0.00
## 4: 6.991000 248.8333 6.233802 0.00
## 5: 6.902333 248.9605 6.574972 0.00
## ---
## 26804: 9.287000 352.6565 4.262119 116.34
## 26805: 9.265000 352.4482 4.243415 93.08
## 26806: 9.233000 350.1066 4.036709 77.56
## 26807: 9.189000 346.2419 3.577402 0.00
## 26808: 9.140000 339.1743 3.634154 0.00
## loc5_RH_2.m.above.ground loc5_TCDC_low.cloud.layer
## 1: 88.30000 0.0000000
## 2: 87.66667 0.0000000
## 3: 87.03333 0.0000000
## 4: 86.40000 0.0000000
## 5: 86.66667 0.3333333
## ---
## 26804: 84.30000 69.8000000
## 26805: 81.30000 69.5000000
## 26806: 86.40000 70.2000000
## 26807: 89.50000 100.0000000
## 26808: 94.00000 99.1000000
## loc5_TMP_2.m.above.ground loc5_wdir_10m loc5_ws_10m loc6_DSWRF_surface
## 1: -4.191000 297.65206 3.913716 0.00
## 2: -4.070333 295.48408 4.243639 0.00
## 3: -3.949667 293.31610 4.573562 0.00
## 4: -3.829000 291.14811 4.903485 0.00
## 5: -3.864333 290.61532 4.996934 0.00
## ---
## 26804: 1.997000 19.05118 4.366314 114.36
## 26805: 2.185000 29.69346 5.556206 91.48
## 26806: 1.463000 47.68443 4.525578 76.24
## 26807: 1.679000 63.83293 3.843080 0.00
## 26808: 1.490000 56.89650 3.948929 0.00
## loc6_RH_2.m.above.ground loc6_TCDC_low.cloud.layer
## 1: 92.10000 28.00000
## 2: 92.50000 25.66667
## 3: 92.90000 23.33333
## 4: 93.30000 21.00000
## 5: 94.03333 21.00000
## ---
## 26804: 73.20000 74.30000
## 26805: 72.20000 79.40000
## 26806: 74.20000 82.60000
## 26807: 89.20000 100.00000
## 26808: 82.50000 100.00000
## loc6_TMP_2.m.above.ground loc6_wdir_10m loc6_ws_10m loc7_DSWRF_surface
## 1: -7.041000 280.4415 1.745607 0.00
## 2: -7.200333 274.3150 1.962468 0.00
## 3: -7.359667 268.1885 2.179329 0.00
## 4: -7.519000 262.0619 2.396190 0.00
## 5: -7.527667 260.3688 2.522906 0.00
## ---
## 26804: -1.953000 324.6584 2.117349 7.32
## 26805: -1.895000 312.2008 2.332197 5.86
## 26806: -1.907000 331.4124 2.160014 4.88
## 26807: -2.371000 300.8462 2.388677 0.00
## 26808: -1.210000 345.0344 2.563677 0.00
## loc7_RH_2.m.above.ground loc7_TCDC_low.cloud.layer
## 1: 65.80000 0
## 2: 66.40000 0
## 3: 67.00000 0
## 4: 67.60000 0
## 5: 67.83333 0
## ---
## 26804: 94.90000 100
## 26805: 94.80000 100
## 26806: 95.00000 100
## 26807: 95.20000 100
## 26808: 95.20000 100
## loc7_TMP_2.m.above.ground loc7_wdir_10m loc7_ws_10m trnd w_day mon hour
## 1: 7.239000 250.5319 5.861836 1 Sal Oca 0
## 2: 7.153000 250.9196 6.254371 2 Sal Oca 1
## 3: 7.067000 251.3073 6.646906 3 Sal Oca 2
## 4: 6.981000 251.6950 7.039441 4 Sal Oca 3
## 5: 6.962333 252.8526 7.291198 5 Sal Oca 4
## ---
## 26804: 3.577000 192.1873 6.417472 26804 Cum Oca 19
## 26805: 2.815000 181.7018 6.656326 26805 Cum Oca 20
## 26806: 2.743000 185.1566 6.047797 26806 Cum Oca 21
## 26807: 2.399000 184.1567 6.081236 26807 Cum Oca 22
## 26808: 2.180000 182.7795 6.641093 26808 Cum Oca 23
## Date actual predicted prevday prevweek residual
## 1: 2019-01-01 -1934.795 -27.562135 287.9239 287.9239 -1900.6192
## 2: 2019-01-01 -750.866 78.071304 287.9239 287.9239 -822.3306
## 3: 2019-01-01 -982.767 48.444963 287.9239 287.9239 -1025.7931
## 4: 2019-01-01 -764.620 29.445657 287.9239 287.9239 -788.4248
## 5: 2019-01-01 -1233.289 2.931106 287.9239 287.9239 -1230.9230
## ---
## 26804: 2022-01-21 0.000 863.904080 2043.4290 -425.3450 -868.6487
## 26805: 2022-01-21 0.000 939.097861 2171.4590 -339.5620 -936.3272
## 26806: 2022-01-21 0.000 1246.179931 2981.0760 -356.4620 -1241.2896
## 26807: 2022-01-21 0.000 1097.096550 2627.5980 -522.6550 -1090.9664
## 26808: 2022-01-21 0.000 269.038245 0.0000 147.8350 -254.2863
## loc7tmp2ag pv loc7tmp2ag2 loc5tmp2ag pd workday loc4tmp2ag whour
## 1: 0 0 0 0 0 1 1 1
## 2: 0 0 0 0 0 1 1 1
## 3: 0 0 0 0 0 1 1 1
## 4: 0 0 0 0 0 1 1 1
## 5: 0 0 0 0 0 1 1 1
## ---
## 26804: 0 0 0 0 0 1 0 1
## 26805: 0 0 0 0 0 1 0 1
## 26806: 0 0 0 0 1 1 0 1
## 26807: 0 0 0 0 1 1 0 1
## 26808: 0 0 0 0 0 1 0 1
## loc3wdir loc6tmp2ag loc7ws loc1wdir loc4wdir loc7wdir loc3tmp2ag
## 1: 0 0 0 1 0 1 0
## 2: 0 0 0 1 0 1 0
## 3: 0 0 0 1 0 1 0
## 4: 0 0 0 1 0 1 0
## 5: 0 0 0 1 0 1 0
## ---
## 26804: 1 0 0 0 0 0 0
## 26805: 1 0 0 1 0 0 0
## 26806: 1 0 0 1 0 0 0
## 26807: 1 0 0 0 0 0 0
## 26808: 1 0 0 0 0 0 0
## loc3tcdc mon1 loc4rh loc4ws loc6rh loc1tmp2ag loc5ws loc2tmp2ag loc7rh2
## 1: 0 1 1 0 0 0 0 0 0
## 2: 0 1 1 0 0 0 0 0 0
## 3: 0 1 1 0 0 0 0 0 0
## 4: 0 1 1 0 0 0 0 0 0
## 5: 0 1 1 0 0 0 0 0 0
## ---
## 26804: 0 1 0 0 0 0 0 0 0
## 26805: 0 1 0 0 0 0 0 0 0
## 26806: 0 1 0 0 0 0 0 0 0
## 26807: 0 1 0 0 0 0 0 0 0
## 26808: 0 1 0 0 0 0 0 0 0
## loc5rh loc1ws loc2dswrf loc6tcdc loc1dswrf loc2rh2 loc3ws loc1tcdc
## 1: 0 0 1 0 0 0 0 0
## 2: 0 0 1 0 0 0 0 0
## 3: 0 0 1 0 0 0 0 0
## 4: 0 0 1 0 0 0 0 0
## 5: 0 0 1 0 0 0 0 0
## ---
## 26804: 0 0 1 0 0 0 0 1
## 26805: 0 0 1 0 0 0 0 1
## 26806: 0 0 1 0 0 0 0 1
## 26807: 0 0 1 1 0 0 0 1
## 26808: 0 0 1 1 0 0 0 1
## loc5wdir loc3rh loc5tcdc loc1rh residuals13 predicted_diretion
## 1: 0 0 1 1 -1907.2329 Neutral
## 2: 0 0 1 1 -828.9373 Positive
## 3: 0 0 1 1 -1031.2120 Neutral
## 4: 0 0 1 1 -794.0657 Neutral
## 5: 0 0 0 1 -1236.2201 Neutral
## ---
## 26804: 1 0 0 1 -863.9041 Positive
## 26805: 1 0 0 1 -939.0979 Positive
## 26806: 1 0 0 1 -1246.1799 Positive
## 26807: 1 0 0 1 -1097.0965 Positive
## 26808: 1 0 0 1 -269.0382 Positivesubmission = tmp2[,list(Date,hour,system_direction,predicted_diretion)]
tail(submission,12)## Date hour system_direction predicted_diretion
## 1: 2022-01-21 12 Positive
## 2: 2022-01-21 13 Positive
## 3: 2022-01-21 14 Positive
## 4: 2022-01-21 15 Positive
## 5: 2022-01-21 16 Positive
## 6: 2022-01-21 17 Positive
## 7: 2022-01-21 18 Positive
## 8: 2022-01-21 19 Positive
## 9: 2022-01-21 20 Positive
## 10: 2022-01-21 21 Positive
## 11: 2022-01-21 22 Positive
## 12: 2022-01-21 23 Positive